A review of “The Knowledge Machine”

The Knowledge Machine: How Irrationality Created Modern Science by Michael Strevens

Science’s irrational origins
Itai Yanai, Martin J. Lercher

Science 2021

What is the scientific method, and what makes it the most efficient approach for generating insight? In The Knowledge Machine, Michael Strevens argues that to answer this question, we must acknowledge the role played by the undisciplined and emotional nature of the humans who carry it out. The book takes readers on a whirlwind tour through the history of science, rendering Arthur Eddington, Louis Pasteur, G. G. Simpson, Lord Kelvin, and many others as “warm-blooded organisms, whose enthusiasms, hopes, and fears mold their thinking far below the threshold of awareness.”

When asked what science is and how it functions, researchers offer a range of conflicting responses, notes Strevens. “Some scientists say that the essence of science is controlled or repeatable experiment, forgetting that experiments are of relatively little importance in cosmology or evolutionary biology. Some say advanced mathematical techniques are crucial, forgetting that the discoverers of genetics, for example, had no use for sophisticated math.”

Strevens argues that an objective scientific method cannot exist, as all predictions from hypotheses rely on auxiliary assumptions such as the functioning of instruments, whose reliability must be evaluated subjectively. He proposes that the distinguishing feature of science is a procedural agreement, which he refers to as the “iron rule of explanation.” This rule holds that differences in scientific opinion must be settled by empirical testing alone. Thus, a scientist cannot argue for one hypothesis over another because it is more beautiful or more appealing philosophically or because it is better aligned with “God’s plan.” The iron rule applies only to official communications. Outside of such venues, scientists may think and believe as they wish.

That only data are capable of formally supporting a hypothesis may seem obvious, yet Strevens suggests that such an approach is inherently illogical. Imagine, for example, suggesting to Aristotle that he should restrict himself to data when arguing in favor of a particular theory. He would have pitied your ignorance. What better support for a theory could there be than an elegant chain of philosophical arguments?

Strevens argues that modern science owes its success to the relinquishing of deep philosophical understanding in favor of the shallow power to predict empirical observations. As Isaac Newton—whom Strevens sees as the first truly modern scientist—wrote: “I have not as yet been able to deduce from phenomena the reason for these properties of gravity, and I do not feign hypotheses. […] It is enough that gravity really exists and acts according to the laws that we have set forth” (1).

Strevens proposes that scientists reason differently in public discourse and private venues. By drawing a clear distinction between formal scientific arguments and informal, behind-the-scenes scientific work, he provides a coherent framework for the divergent ideas of earlier philosophers of science: Karl Popper’s ideas on the falsification of hypotheses (2) form the basis of formal scientific discourse; Paul Feyerabend’s observations highlight the subjectivity of daily work, including the evaluation of assumptions (3); and the apparent security of a scientific paradigm guided by the iron rule compels scientists to perform elaborate experiments, thus generating data of otherwise unimaginable quantity and detail—a phenomenon described by Thomas Kuhn (4).

Strevens frames the toiling life of data generation as the cost scientists pay to gain access to the sacred halls of scientific excellence. What he overlooks is the supreme “pleasure of finding things out” (5). In his autobiography, French biologist François Jacob proposed the notion of “night science,” in which scientists generate new ideas and hypotheses in often unstructured thought processes (6). This approach, he argued, complements “day science,” wherein new ideas are tested empirically and reported formally. Thinkers such as Aristotle perceived day and night science as intertwined in a single process. Newton and his contemporaries founded modern science by separating them into distinct undertakings. While Strevens’s iron rule may indeed be the foundation of modern science’s success, the methods scientists use to come up with new ideas remain elusive.

References and Notes:
1. I. Newton, The Mathematical Principles of Natural Philosophy (Benjamin Motte, 1687).
2. K. Popper, The Logic of Scientific Discovery (Hutchinson, 1959).
3. P. Feyerabend, Against Method (Verso Books, 1975).
4. T. Kuhn, The Structure of Scientific Revolutions (Univ. of Chicago Press, 1962).
5. R. Feynman, The Pleasure of Finding Things Out (Perseus Books, 1999).
6. F. Jacob, The Statue Within (Cold Spring Harbor Laboratory Press, 1995).

About the author

Yanai is at the Institute for Computational Medicine, NYU Langone Health, New York, NY 10016, USA. Lercher is at the Department of Biology and Institute for Computer Science, Heinrich Heine University, 40225 Düsseldorf, Germany.


Leave a Reply