Event Report - The Success of Failure: Perspectives from the Arts, Sciences, Humanities, Education, and Law

December 7-8, 2017
Cowin Auditorium, 147 Horace Mann Hall, Teachers College, Columbia University

For additional details and quotations from each session, please read the Extended Failure Report

  • Speakers: 
    • Harry Collins (Social Sciences; Cardiff University)
    • David Kaiser (History of Science and Physics; Massachusetts Institute of Technology)
    • Jutta Schickore (History of Science; Indiana University Bloomington)
  • Respondents: 
  • Chair: Pamela Smith (History; Columbia University)

This panel drew on historical examples to illustrate that “success” and “failure” are often determined by extenuating circumstances such as inadequate technology and social bias rather than objective and conclusive evidence. The speakers argued that when determining whether or not a past outcome was a failure, one must consider the tools and knowledge of the day and not just the results. In fact, results that were deemed personal failures at the time often inspired greater work that eventually had positive outcomes for society. Harry Collins gave the example of how Joseph Weber’s failed attempt to detect gravitational waves (which disparaged his career) later inspired Weiss, Barish, and Thorne to take up his mantle. Over time, the technology needed to successfully prove Weber’s hypotheses was gradually improved and they were able to detect gravitational waves. They won the 2017 Nobel Prize in Physics and credited Weber’s pioneering work. Jutta Schikore took an example from the history of bacteriology and described Ernst Hallier’s work on the study of fermentation and bacterial cultures. At the time, the lack of technology needed to create a serious science of bacteriology limited the success of his experiments; however, his failed and discredited work eventually inspired early studies of wound infection.

David Kaiser looked at the role of social bias in determining success. He gave the example of Jan Hendrik Schön, the German physicist who falsified data from his experiments but was nevertheless well received by the scientific community because he had produced results that the scientific community hoped and expected to find. Alondra Nelson brought up the idea of science as an exercise of power in which an issue like climate change can either be considered a matter of scientific consensus or a matter of debate depending on the social and political context of a given point in time. The resounding conclusion was that success and failure are mutable and time is often needed to truly determine if an outcome was a success or failure.

View the "Failures in Science From the Perspective of History" session video

This panel looked at failures in regulations that restrict behaviors that directly threaten public health, safety, and well being. Similar to the previous panel, the speakers illustrated how perceptions of “failure” and “success” at any given time are mutable, subjective and largely dependent on extenuating circumstances such as the moral climate of the day and the values of the people who are in a position to determine and foster the success of a law or initiative. Tim Wu made an insightful point that while lawbreaking is typically viewed as negative and therefore a “failure,” in certain cases where laws are seen as unjust, law-breaking can be seen as a form of civil disobedience, an act that is viewed as positive and therefore “successful.” He continued to illustrate the role of morality in law breaking with the issue of online fantasy sports. Despite being technically prohibited in New York State by the constitution, Wu described the extent to which online fantasy sports leagues thrive, subverting what is seen by many as illogical and outdated laws against gambling. What was once considered a social taboo and a legal offense is now viewed in most spheres as a harmless vice.

Carl Hart noted the role that social bias plays in determining what is considered morally bad (a failure) or morally good (a success) with the example of drug policy in the United States. He argued that U.S. drug policy has historically not been based on pharmacology but sociopolitical biases based on race and class. Scientific research proves that drugs and drug use are not as bad as the government makes them out to be, however, vilifying them is a way for the government to subjugate and criminalize the people–mostly people of color–who use them. The real failure is not in using drugs but in not adequately connecting scientific research with the law. If they were aligned, we would not have the illogical and racist drug policies that we have today that primarily target people of color.

Similarly, Alondra Nelson discussed the role that political morality plays in determining the success or failure of government initiatives and talked about the transformation of the Office of Science and Technology Policy (OSTP) following the end of the Obama administration. The successes of science policy under the Obama administration recognized the long-term benefits of science as a strategy. Yet since the election of President Trump, massive departures of staff and the dismantling of the OSTP website have brought a close to this hallmark of the Obama presidency. The wiping of the EPA website of all its data and reports is representative of the diminished value of expertise and facts, signaling the demise of a bully pulpit role for public science.

This session looked at science writing and the kind of studies that ultimately get published and recognized in scientific journals. Jutta Schikore, a session respondent, noted that in the 18th- and early-19th- centuries, official science writing saw the value in publishing writing that looked at successful results, unsuccessful results, and methodology. It was only with the emergence of mass academic publishing that the standardization of form and scientific writing began to omit details and focus only on features of the research that were novel, innovative and “successful”.

John Spiro supported this argument by noting that while some recognized journals (such as Nature) value studies that look at what has not worked in research, this is not the case with scientific journals more broadly. He posits that this could be a result of, among other things, social sensitivity around publically refuting the work of other colleagues. To address this issue, he advocated the idea of “pre-prints,” unofficial sites for scientific articles that do not have to meet the criteria of novelty or orient around breakthroughs but can focus on studying methodology even if the results were not successful. He argues that a culture of publishing results regardless of outcome would lead to more dialogue and acknowledgment of process and methodology which are in fact crucial to producing successful results.

This session focused on the importance of valuing and emphasizing process rather than results in teaching and setting people up for success, whether it be in a traditional classroom setting or empowering welfare recipients in a social services agency. Xiaodong Lin spoke to this point in the context of her work with grade school students studying science. She stated that the cultural association of the Science, Technology, Engineering, and Math (STEM) field with “geniuses” (Albert Einstein, Marie Curie, Thomas Edison, etc.) makes STEM careers appear untouchable and as a result might dissuade students from entering such fields. To counter this, she suggested providing detailed stories of actions and strategies that these scientists took throughout their careers to overcome obstacles. The point being to make these figures more relatable, offer a fuller and more human picture of scientific research, minimize the stigma around failure, and encourage perseverance. Introducing an element of humanity into the sciences allows students to see that failure functions in various ways and to take a healthier view of obstacles to achievement.

Similarly, Luz Santana noted that the assumptions made about people on welfare (that they are failures, inherently dependent and incapable of taking care of themselves etc.) can have a detrimental impact on the ability of welfare recipients to develop the skills necessary to get off welfare. She stated that in addition to providing services and taking care of immediate needs, welfare offices need to be able to envision welfare recipients as capable of success and teach and empower them to advocate for themselves so that they can actively navigate the systems of politics and governance that have had a hand in their circumstances. This approach would ultimately lead to a more sustainable change in their lives than simply receiving a handout.

View the "Importance of Teaching Failure in Education" session video

This panel looked at narratives of success and failure. Jennifer Hecht posited that doubt and uncertainty drive the pursuit of knowledge and illustrated her argument with examples from her research on the history of atheism and enlightenment science. Scott Sandage explored the categories of definition and identification linked to success and failure in the current political discourse. Drawing on his book Born Losers: A history of failure of America, Sandage examined the place and poetics of failure in a culture defined in many ways by its relationship to success as a governing ideal. Scott Sandage also explored his own fear of failure, tracing the language of failure and losing in American idiom.

  • Speakers: 
    • John Black (Telecommunications and Education; Teachers College, Columbia University)
    • Catherine Chase (Cognitive Studies; Teachers College, Columbia University) 
    • Robert Siegler (Cognitive Psychology; Carnegie Mellon University)
    • Lisa Son (Psychology; Barnard College, Columbia University)
  • Respondents: 
    • Carl Hart (Psychology; Columbia University)
    • Mario Livio (Astrophysics; University of Nevada, Las Vegas)
  • Chair: Xiaodong Lin (Cognitive Studies; Teachers College, Columbia University)

This session looked at how people’s mindsets regarding effort and time impact if a goal will be pursued and ultimately achieved. Robert Siegler argued that failure is beneficial because it signals to us that we should stop wasting our time and focus our effort on other things that we do well. He asked the audience to consider several things such as how much effort someone should put into an unsuccessful endeavor before deciding it is time to quit and if there is merit in an educational system that has students struggle in a mandatory subject in which they have no skill or interest. How much effort, in fact, should be put into an endeavor?

In response, John Black noted in his study of gaming, that there is a high correlation among gamers between what they deemed are winning strategies for a game and which strategies require the least amount of effort. Exploring the psychology of risk, Catherine Chase outlined ways in which people’s openness to an iterative design mindset which risks mistakes and embraces failure allows people to gain the vital early feedback necessary to increase chances for success. Similarly, Lisa Son stressed the importance of understanding the structure and timeline of a learning curve and thinking about what determines when we should stop putting effort into achieving a goal.

View the "Psychology of Failure" session video.

This session examined the art of science and the roles that curiosity, iterative process, and feedback play in how science is conducted. Mario Livio started the session by looking at the most basic driver of scientific advancement–curiosity. Two scientists who were driven by what Livio termed epistemic curiosity (a pro-active curiosity that is specific and focused on a determined topic) were Leonardo Da Vinci and Richard Feynman. Both worked in a variety of fields and their insatiable curiosity was driven mostly likely in part to the fact that epistemic curiosity is associated with the area of the brain that is linked to the anticipation of award. The idea of curiosity is also the premise of what Peter Norvig termed the “test-driven development” system that is used in the creation of software. Like the design-thinking methodology described by Catherine Chase in a previous panel, “test-driven development” is driven by exploration and the idea of pivots so that failure is reconceptualized not as a stumbling block but as a point at which to change direction and explore something new. He reminded the audience that Flickr, the popular photo-sharing platform, was a direct result of the Dot Com Crash – demonstrating how identifying and shepherding successful parts of larger failures can foster success. Norvig did ask the audience to consider the limits of curiosity.

Exploring and pursuing the most successful iteration of a product over a less successful version is desirable (and as Livio pointed out–probably linked to a pleasure center in our brain) but testing 41 shades of blue for a new product does not always make sense from the perspective of efficiency. One must have a sense of when there has been enough feedback and complete a project. Michael Shadlen noted that the notion of formulating ideas along the way and embracing uncertainty (as in the “test-driven development” model, involves moving outside of one’s comfort zone that can be at once liberating and daunting. In such iterative processes, self-reflection becomes a tool that scientists can use in the laboratory to look for ways to learn from their mistakes and prevent them from happening again in the future.

This panel examined the relationship between individual and collective gain. Katharina Pistor looked at this relationship in the context of state and international laws governing business and economics. She noted that while Adam Smith argued that entrepreneurs will always invest their wealth in their local economy in order to support their countrymen, globalization has challenged this assumption by making it easier for business to be located anywhere in the world. The failure of States to keep wealth within their borders poses a question: is it possible for States to reverse the liberalization of individuals within markets and restore the link between society and individual gains? Should there be limits to globalization?

Josh Wolfe also looked at the dynamic between the individual and the collective but in the context of how investment decisions are made. He noted that failure in business comes from a “failure to imagine failure.” Understanding where there are opportunities for improvement and growth allows venture capitalists to identify and target potential areas for investment. The ability to think and act independently and see potential where others cannot is key to pursuing ideas that might be risky but have the potential to see a higher return on investment. Despite this, he noted that individual investors still succumb to strong social pressure to look at where the collective is investing so that they can achieve safe and predictable outcomes.

This session looked at failure as a key component of the creative process. Chris Washburne argued that failure is a central–even beautiful–component of jazz music as a genre. He noted that improvisational jazz is full of “mistakes” such as mishearing and mistiming. Although barely perceptible to audiences, to a trained ear they are in many ways the most exciting part of the performance. He noted when Herbie Hancock and Miles Davis were playing together and Hancock’s mistake on the piano forced Davis to adapt to adversity and create an unexpected harmony. John Collins and Sara Jane Bailes discussed failure as a recuperative condition that insights us to go on in the face of adversity.