The Golem: What You Should Know About Science (Summary)
For years, physicists tried to build a working TEA-laser based on published scientific papers, and they all failed. The blueprints were perfect, the math was sound, but the lasers wouldn't fire. It was only when they visited the original lab that they discovered the secret ingredient: the unwritten, 'folk' knowledge of the technicians, like which specific brand of epoxy to use and how to 'feel' for the right vacuum seal. The 'objective' knowledge of science was useless without the messy, human craft.
The Vicious Circle of Scientific 'Proof'
To know if an experiment is working, you need to know the 'right' answer it should produce. But how can you know the right answer without a working experiment? This paradox, called the 'Experimenter's Regress,' reveals that scientific consensus often relies on social trust and negotiation, not just cold, hard facts.
In the 1970s, Joseph Weber claimed to have detected gravitational waves. Other labs couldn't replicate his results. Was Weber's machine uniquely sensitive, or was it flawed? Was everyone else incompetent? With no independent way to 'see' a gravitational wave, the scientific community had to decide who to believe and what counted as a 'good' experiment, a debate that lasted for years.
Certainty Is an Illusion Created in Hindsight
Textbooks present scientific discoveries as a clean, linear path. In reality, at the moment of discovery, the data is messy, controversies rage, and the 'correct' interpretation is far from obvious. Scientific facts are constructed, not simply found.
During the solar neutrino puzzle, Ray Davis's experiments consistently found only one-third of the neutrinos predicted by established solar theory. For nearly 30 years, the community debated: Was Davis's complex, difficult experiment wrong, or was the fundamental theory of how the sun works flawed? The certainty we have now—that the theory was incomplete—was forged from decades of uncertainty and messy debate.
Science is a Craft, Not a Recipe
Much of scientific expertise is 'tacit knowledge'—the unwritten, intuitive skills and know-how that can't be learned from a textbook. Replicating an experiment is often impossible without direct, hands-on apprenticeship.
The study of memory transfer in planarian worms in the 1960s serves as a key example. Some labs could replicate the finding that 'memories' could be transferred by feeding trained worms to untrained ones, while others failed completely. The difference came down to subtle, unspecifiable skills in handling and observing the worms—a craft that some researchers had and others simply could not acquire from the literature.
The Border Between Science and Pseudoscience is Negotiated
The line separating 'real science' from 'fringe ideas' isn't fixed or logical; it's a social boundary drawn by the scientific community. What is considered crackpot science in one era can become a legitimate field of inquiry in another, and vice-versa.
The theory of cold fusion, announced in 1989 by Martin Fleischmann and Stanley Pons, is a prime case. Initially hailed as a revolutionary discovery, it was quickly relegated to the status of pathological science after widespread failures to replicate it. The boundary was drawn swiftly and decisively, not just on evidence alone, but on the social and institutional reaction of the physics community.
