364 | Stuart Firestein on How Science Relies on Ignorance and Failure

/ Philosophy / 11 Comments

One of the paradoxes of science is that it seeks objectively true understanding of the world, but its methodology is driven by ignorance, failure, and uncertainty. Some phenomena we understand pretty well, but interesting research happens at the boundary of what we do and don't know. And there is no foolproof algorithm for moving in the right direction; we need to make conjectures and test them against the world. Biologist Stuart Firestein has been advocating for a better public understanding of the true methods of science, most recently in his new book It Could Be Otherwise: Science In the Age of Uncertainty.

Stop piecing your software together. Go to Odoo.com/mindscape to learn more. #ad

Mindscape listeners can try StatsKey Pro free for one month with code MINDSCAPE. After that, it’s $4.99 per month unless canceled. #ad

stuart firestein

Support Mindscape on Patreon(opens in new tab).

Stuart Firestein received his Ph.D. in neurobiology from the University of California, Berkeley. He is currently Professor of Neuroscience at in the Department of Biological Sciences at Columbia University, and Fractal Faculty at the Santa Fe Institute. He is a fellow of the American Association for the Advancement of Science, the Sloan Foundation, and the Guggenheim Foundation. His research studies cellular mechanisms of signal transduction and olfaction. His previous books include Ignorance: How It Drives Science and Failure: Why Science Is So Successful.

11 thoughts on “364 | Stuart Firestein on How Science Relies on Ignorance and Failure”

  1. Bernard

    Excellent, thank you. As a science educator (unprofessional, unpaid) I found this most insightful. You may find this quotation from James Clerk Maxwell to be apposite: “Thoroughly conscious ignorance is the prelude to every real advance in science.”

  2. Cleon Teunissen

    I think the story of Helicobacter Pylori is very instructive.

    From history of medicine:

    For microorganism caused diseases there was, from around 1880 to around 1900, in succession identification of the microorganism causing the disease, from cholera to whooping cough, there were many. Around 1900 that streak ended. In 1983 Helicobacter Pylori was formally described as the main causal agent of gastric ulcers.

    To the question why it took so long.
    It would appear that a window of opportunity had closed.

    Part of the explanation: you would not expect that a bacterium would be able to multiply in the acidic stomach environment. That expectation had shifted towards dogma. The medical community had come to attribute gastric ulcers to lifestyle factors. With that expectation pattern widespread: opportunity to recognize the actual main causative agent went away.

    This goes to psychology of science.
    The default mindset of the human brain is to interpret new information in terms of existing expectation. Existing beliefs are active agents, modifying perception.

    Once a window of opportunity has closed: as an expectation is in place longer and longer the corresponding bias becomes less and less likely to be perceived.

    As pointed out in the podcast: there is no algorithm.
    From within the expectation pattern there isn’t a way to ascertain whether in the past something was overlooked, and if something was, what the overlooked circumstance is.

  3. Philip Koop

    I am disappointed that you let Firestein pass when he said that learning facts about science is like learning history; that is nothing at all like the practice of history.

    I think that Timothy Snyder’s typology of sources would be illuminating here. As he puts it, the distinction between a primary source and a secondary source lies in the intentions of the reader, it is not an inherent property of the text. When I read a text for the sake of the information its author intended to convey to me, then I am reading it as a secondary text. But when I read it “against the grain” in order to learn something about the author, or the society the author lived in, or just to divine the facts behind the author’s claims, then I am reading the same text as a primary source. Which is to say, the epistemology that eschews certainty and totality in favour of credences and piecewise explanations is not specific to science, it permeates all of modern society. You use it even when making the prosaic decision of which teapot to purchase, as Ada Palmer once said.

    I doubt that you could get Snyder on the podcost, but it sure seems that you, and perhaps your listeners, would profit by interviewing *some* historian about the meaning and practice of history.

  5. dmf

    thanks for this we need more sociology of science (in the STS mode) to go with all the philosophy of science.
    Someone who does both well (as a well as having a Phd in physics) is Andrew Pickering who would be a great guest to have on to talk about The Mangle of Practice.

  6. Dr. Sonali Sengupta

    Perspective from a single biologist.
    I being a molecular biologist would agree with the ignorance part ‘we do not know’ or what we know is the drop but not the ocean to put it poetically. I tend to disagree, as I believe assumptions & falsifiability ( proving theories to be false and assuming what is remaining to be true) are key factors driving progress in the field at least in molecular biology. Paraphrasing John Gurdon , Nobel laureate biologist, ‘failure is a part of the process’, which is very different from ‘science relying on failure’. Sometimes experiments work at the first attempt if assumptions are appropriate. Scientific method is a layered process.

  7. Dr. Sonali Sengupta

    Perspective from a single biologist.
    I being a molecular biologist would agree with the ignorance part ‘we do not know’ or what we know is the drop but what we do not know is the ocean to put it poetically. I tend to disagree, as I believe assumptions & falsifiability ( proving theories to be false and assuming what is remaining to be true) are key factors driving progress in the field at least in molecular biology. Paraphrasing John Gurdon , Nobel laureate biologist, ‘failure is a part of the process’, which is very different from ‘science relying on failure’. Sometimes experiments work at the first attempt if assumptions are appropriate. Scientific method is a layered process.

  8. Mark Ikeda

    From the classroom observations of a retired community college professor of Biology; I think that the public’s default example of a process of understanding objective reality come from, unfortunately, methods associated with debate in political/legal realms. For example analyzing students’ disagreement with the scientific notion of evolution becomes surprising (to them) when a question is posed about the components they have issue with; heritability?, differential reproductive success?, etc. The idea of science as a bottom up process based upon objectively gathered data is at odds with the top down notion of starting with a statement and gathering an array of information in support of that statement.

  9. David Vognar

    See Thomas Kuhn, On the Structure of Scientific Revolutions. Would you like to come to Oak Lawn and talk rather more fulsomely than obliquely? It can be a quaint and nice place at times.

  10. David Vognar

    Perhaps the history of the errors of mankind, all things considered, is more valuable and interesting than that of their discoveries. Truth is uniform and narrow; it constantly exists, and does not seem to require so much an active energy, as a passive aptitude of the soul in order to encounter it. But error is endlessly diversified; it has no reality, but is the pure and simple creation of the mind that invents it. In this field the soul has room enough to expand herself, to display all her boundless faculties, and all her beautiful and interesting extravagancies and absurdities.

    Benjamin Franklin

  11. Neil Greenspan

    The Mindscape episode with Stuart Firestein was not particularly informative. Firestein’s comments were frequently excessively sweeping and not clearly based on actual evidence. For example, does he really believe that no one teaching undergraduate science in the U.S. is aware of the nuances and complexities of scientific methods.

    My own experience belies this lazy generalization. Over 50 years ago, I was an undergraduate science student. Also in those years, I took philosophy of science and logic and incorporated concepts from these and other areas of philosophy in my subsequent scientific thinking as a graduate student, medical student, postdoc, physician and basic biomedical researcher at various academic medical centers as well as in my clinical activities and teaching. My experience was not unique. I think it is likely that at least some of my faculty colleagues exhibit some awareness of the importance of the various, sometimes non-canonical, pathways by which scientific insights are generated and share this knowledge with students and trainees.

    Professor Firestein is right that overall improvement in how science is taught is to be desired, but his messages are not especially new. Peter Medawar, in his 1968 Jayne Lectures, published by the American Philosophical Society in 1969 as “Induction and Intuition in Scientific Thought”, addressed the ways that research publications failed to reflect the actual messiness of scientific advance. There are of course many other publications of various sorts that discuss the need to learn from failure or that document that it is sometimes better to be naïve about the beliefs or hidden assumptions in a given subfield of scientific investigation, but that does not mean investigators should desire failure or cultivate ignorance.

    Some failures are dead ends or worse, diverting a field into unproductive directions of inquiry. Ignorance can likewise contribute to unproductive experiments and the waste of limited resources.

    The most significant exemplification of Professor Firestein’s failure in scientific communication, but not the only one, comes in the form of his ill-advised comments about the workings of biological evolution. Mutations occur randomly in a highly technical sense, which is that genetic mutation frequency is not related to functional impact. In other respects, mutation is not necessarily random as it is constrained by the biophysical properties of the genetic material and the relevant biochemical processes. For example, at a given genomic site, the probabilities of a given nucleotide being replaced by any of the three alternative nucleotides are not necessarily equal depending on the mutational process involved. In this instance, Firestein is not completely wrong but is well short of maximally clear.

    More consequential and contrary to Firestein’s claim, natural selection is not “kind of random,” which is an astonishing instance of ignorance in a former chair of a biology department. The critical feature of selection, in fact, is that it is the utter antithesis of randomness. I do not see how this example of ignorance or surprising sloppiness in thinking advances science, and therefore, it is an example of unfortunate failure with no upside. The host should not have allowed such a misleading claim to pass unchallenged.

    I agree that the combination of mutation and natural selection do create, in concert with other factors, the amazing diversity and adaptation observed in the living world. Of course, this insight is not new with Professor Firestein and is only partly but definitely not completely attributable to random processes. Without the highly non-random shaping of natural selection, random mutation would obscure the exquisite adaptations we observe.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top