What this covers

Sean Carroll talks with the neuroscientist Stuart Firestein, who studies olfaction at Columbia, about his book It Could Be Otherwise. The conversation runs from how experiments work to how science is taught. Firestein says a confirmed hypothesis yields a measurement while a refuted one yields a discovery, and that labs need both kinds of student. He also notes that the Hypothetico-Deductive Method says nothing about the imaginative step of generating a hypothesis. Carroll raises how lay audiences hear probability, as zero, one half, or one, so "probably true" sounds like either ignorance or certainty.

The range is wide. Firestein argues that serendipity matters only to people already working on a problem and willing to keep counterintuitive results. He points out that magicians exploit Occam's Razor, and that Voltaire coined "optimism" only in 1759. He reads Peirce and John Dewey as champions of uncertainty, with Peirce anticipating chaos theory.

The more contested claims come later. Firestein ranks Darwin's account of order arising from randomness above Einstein or quantum physics as a shift in outlook. He treats probability as intrinsic to the universe rather than a gap in knowledge, and describes reality as dappled, with islands of predictability amid irreducible uncertainty. He defends pluralism without relativism, drawing on Hasok Chang's idea of operationally coherent activity. Much knowledge, he says, is provisional truth. Carroll pushes back by arguing that a unified theory of everything remains meaningful in principle. Both tie biology's mix of contingency and convergence to the Santa Fe Institute's edge between order and chaos.

Sharpest takeaway

Firestein argues that science is driven by structured ignorance, failure, and provisional knowledge rather than accumulated facts or a fixed method, and that public misunderstanding of this process, reinforced by how science is taught, undermines trust in expertise.

  • Science is organized ignorance: well-defined unknowns generate new questions, and failed experiments reveal what we did not know we did not know.
  • No algorithm or method reliably produces discoveries or judges which research direction will pay off; hypotheses are cheap and counterintuitive thinking is essential.
  • Science's distinguishing feature is openness to uncertainty; pseudoscience claims certainty, and an effective science education should teach asking questions, not memorizing facts.

The argument · threads4 threads · 76 claims
0.85

Science is a structured, bounded pursuit of the unknown, not facts or random guessing.

4 pointscentrality 5/5
  • Science is not about facts alone; it is a structured pursuit of what we do not yet know, where the unknown is bounded and shaped by what we already know, so the process is not random guessing.

    “We know a lot, there's a lot that we don't know, and that structured ignorance that we're trying to uncover the facts hidden amongst is what science is all about.”

  • Science is taught as a set of facts to memorize and be tested on, rather than as the process of doing science, which leads to a public misunderstanding of how science works.

    “we sort of teach the product of science and not the process of science”

  • Even if a unified theory of everything may never be reached in practice, it remains meaningful to hold that in principle one exists, and science should not treat unreached ultimate truth as its only target.

    “in principle, there might be a single unified theory of everything, right? And in practice, we might never get there”

  • Carroll's rule that when a question begins 'Is it possible that...' the answer is always yes, which Firestein's book is meant to justify because the space of possibilities beyond current knowledge is large.

    “if they ever ask me a question that begins with, 'Is it possible that?' The answer is always yes”

0.71

Public misreading of uncertainty and expertise makes candor and healthy skepticism hard to balance.

4 pointscentrality 4/5
  • Lay audiences tend to treat probability as having only three values, zero, one half, or one, so a scientist who says something is probably true but uncertain is heard as either knowing nothing or being certain, while the 80 percent range is heavily populated in real science.

    “they think that the only probabilities are 0, 0.5, and 1. So if you say, well, we think that something is probably true, but we're not sure, they either read that as we have no idea or we are sure.”

  • The Babe Ruth conspiracy claim shows how hard it is to balance skepticism with respect for evidence, since doubt that ignores abundant evidence or builds complicated theories is not the same as healthy skepticism, and Occam's razor still applies.

    “it just drives home how really difficult it is to express and convey that line between being skeptical, being doubtful, and yet not ignoring the actual evidence”

  • Many people form opinions out of insecurity, assuming they know better than experts, and this drives resistance to expert judgment that is hard to fix.

    “that sort of insecurity that ends up being the driver for so many people's opinions”

  • Public health officials during COVID-19 may have been overly cautious in disclosing their uncertainties, and balancing candor about uncertainty with public trust is a very tough line to draw.

    “maybe you could argue that they were a little bit overly cautious in letting the public know about their uncertainties. That's a really tough line to draw.”

0.43

Institutions favor mainstream theories, and sincere minority voices are hard to tell apart from crackpots.

2 pointscentrality 3/5
  • A practical test for crackpot versus sincere messages is that sincere correspondents ask questions and engage with the recipient, while crackpots talk at you and show no interest in the response.

    “The sincere people ask questions. The crackpots just want to talk at you”

  • Departments hire the researcher on the 90 percent favored theory rather than the 10 percent minority theory, so no institution systematically supports plausible but minority research pursuits.

    “And 100% of the physics departments are going to hire that person, right? So it's not like 10% of them are pursuing theory two.”

0.36

Nature's indifference and Darwin's impact shape how humanity understands its place in the universe.

3 pointscentrality 2/5
  • Biology involves both contingency and convergence, and the coexistence of these views suggests we live at the edge between order and chaos, which both Firestein and Carroll see as a Santa Fe Institute lesson.

    “we live on the edge of in between order and chaos in some very real sense.”

  • Darwin changed humanity's understanding of its place in the universe more than Einstein or Newton did.

    “certainly in terms of our place in the universe as human beings, Darwin did more to affect that than Einstein or Newton did.”

  • Einstein's view that nature is subtle but not malicious is the scientist's hope, in contrast to magicians who are deliberately malicious, though nature has no particular drive either way.

    “Einstein's opinion that nature is subtle but not malicious. Like, the magicians are malicious.”