Skip to main content


A new history acts as a field guide to the fictional imps invoked by scientific thinkers

Bedeviled: A Shadow History of Demons in Science

Jimena Canales
Princeton University Press
416 pp.
Purchase this item now

What role could imaginary creatures possibly play in the most rational, systematic, and discovery-oriented enterprise humanity has ever created? Did the Enlightenment not teach us to dispense with the supernatural and provide us with the power of reason to exorcise demons from modern science?

As Jimena Canales reveals in her thought-provoking and highly readable book Bedeviled, fictional imps that help us explore the limits of what is possible pervade the history of scientific thinking, namely in the form of thought experimentation. Although this is not a new topic of inquiry, Canales comes at it from a refreshingly original perspective, examining the ways that various scientific thinkers throughout the centuries have made use of imaginary demons.

In the book’s 10 chapters, Canales recounts science’s most famous fiends, including Laplace’s demon, a being who could predict every occurrence in the universe, and Maxwell’s demon, who could violate the second law of thermodynamics. More contemporary examples are discussed in the book’s later chapters and include the self-programming electronic circuits that Norbert Wiener referred to as “demons” and the lines of code known as “daemons” that are central to today’s computer communication infrastructure.

At the outset of her scientific demonology, Canales presents a classic example: Descartes’s demon. In his Meditations on First Philosophy (1641), the French philosopher, mathematician, and scientist René Descartes attempted to establish a firm foundation for scientific knowledge. To do so, he used a method known as “Cartesian doubt,” which encouraged a skeptical approach to reality. After arguing that our sensory experience is often erroneous and also that we can never be certain that we are not dreaming, Descartes imagined an even worse scenario: What if a powerful and clever demon systematically presented us with a false reality? If this were the case, we could not trust our senses at all. There is, however, one feature of reality that such a demon would be unable to manipulate. “Cogito, ergo sum,” reasoned Descartes, “I think, therefore I am.” From this foundation, Descartes constructed an approach to scientific knowledge that fundamentally influenced Western philosophy and modern science.

In her chapter on computer daemons, Canales discusses Searle’s demon, an imp introduced by philosopher John Haugeland in response to an assertion made by philosopher John Searle that computers do not understand the symbols they manipulate (1). Searle argued that artificial intelligence is comparable to a native English speaker with no knowledge of the Chinese language, who is locked in a room and asked to provide sensible responses in Chinese to inquiries made in that language. Given an English rule book that instructed the user how to correlate Chinese symbols with appropriate responses, the person could convincingly convey that he understood the messages being relayed without actually understanding them. Similarly, argued Searle, the central processor of a computer translates without understanding any of the symbols it manipulates. But, asked Haugeland, what happens if “a superfast person”—Searle’s demon—took over an individual’s neural processing, replicating that person’s exact responses? What is then causally responsible for the person’s understanding—her neurons or the demon?

We learn a great deal about the scientists who have used demons to confront difficult challenges and approach apparent paradoxes. For instance, along with her description of the article in which Albert Einstein invoked “ghosts” in order to question traditional notions of space (and, later, time), Canales includes details about Einstein’s health problems, his broken marriage, and how Arthur Eddington helped to confirm his theory of general relativity.

The reader also learns about the role that cultural factors played in shaping scientific thought at different times and in different contexts. For example, we learn why Descartes warned against the “dangers” of reading novels like those that intrigued Don Quixote and also about the role played by the Renaissance painter Raphael in Erwin Schrödinger’s conceptions of life and negative entropy.

As Canales maintains, demons are “neither just psychological delusions or simplistic heuristic fictions nor simply auxiliary midwives who help scientists deliver knowledge.” On the contrary, she argues that the notion of a (scientific) demon should be regarded as akin to other, more established philosophical notions or tools, “such as concepts, numbers, classes, and categories”—a welcome contribution to the philosophy of scientific discovery that deserves further scholarly attention.

References and Notes:
1. J. R. Searle, Behav. Brain Sci. 3, 417 (1980).

About the author

The reviewer is at the Department of Philosophy of Religion and Science, Ruhr-Universität Bochum, 44801 Bochum, Germany.