§24. Prehistory of Experiments

It seems credible that Pythagoras (sixth century bce) carried out system­atic experiments to determine the octave. The oldest account of his research, followed by all later authors, is from Nichomachus of Gerasa (second century ce).

There seems no older trace than this of experimental practice, but there­after the evidence picks up. Empedocles thinks experimentally about pneu­matics. Hippocratics record experimental observations on the vacuum. Aristotle studies the reaction of substances to simple tests (boiling, freezing, burning), and the research continued in the Lyceum. Aristotle describes several experiments in his works (whether he performed them is another matter), and we have seen the place of experience in his theory of science. His school continued to welcome experimental work, by Theophrastus, the second head, and following him Strato, who demonstrated the compress­ibility of air, from which he inferred small local vacua, refuting Aristotle’s teaching against the vacuum. Unfortunately nothing of his work survives (unless he is the author of the ps.-Aristotle Mechanics).³

Greeks tended to cast their theories in a form that did not invite exper­imental test. Any role for experience is usually limited to corroborating the author’s view or refuting another, with a tendency to favorable overin­terpretation of evidence.

Philo of Byzantium (ca. 200 bce) describes research on mechanical problems connected with the construction of a catapult, reporting system­atic experimental inquiry by Alexandrian engineers in search of the op­timal power from their war machines. A timber frame housed the twisted skeins that drove the catapult. If the skeins were too large and their force too great, the frame shattered. Wanting to find the greatest elastic force for a feasible scaffold, these engineers used the method of parameter variation, systematically modifying the size of the bore holding the twisted skeins, experimenting on scaled-down prototypes, then replicated the proportions to build a working machine. Philo preserves their formula. The optimal di­ameter of the spring should be 1.1 times the cube root of the stone's weight multiplied by 1,000.

“It was necessary to determine this diameter not accidentally or haphaz­ardly,” he writes, “but by some definite method.... This could not be done except by increasing or decreasing the diameter of the bore and testing the result.” He adds, “It is evident that it is not possible to arrive at a complete solution of the problem involved merely by reason and by methods of me­chanics, and that many discoveries can be made only as a result of a trial.” Experimental parameter variation remains a method in engineering re­search, especially when engineers seek knowledge in areas scientific theory has not illuminated.

Alexandrian mathematician Heron called for the unification of natural philosophy and mechanics (§21). Strato may have done the same thing. But experimentalism never became mainstream in the natural philosophy of an­tiquity. What eventually discredited Aristotle's philosophy of nature was not more experience or greater openness to experience. It was a different kind of experience, an orchestrated experience, arranged in accordance with a hy­pothesis, an experimental experience. The value of such experience begins in the technical arts. Ancient science was lost with failing Roman imperium, but the industrial arts survived and improved in uninterrupted advance over the Romans. This flourishing of the arts produced what historian Lynn White describes as “the chief glory of the middle ages,” which is not the cathedrals or scholastic philosophy. It is for the first time in history to have built a complex civilization that was “upheld not on the sinews of sweating slaves and coolies but primarily by nonhuman power.” The first technological civilization.⁶

From Bede in the seventh century, European scholars favorably noticed the arts and praised improvements. By the twelfth century Europeans are using many technologies unknown to the Romans, and had developed an interest in experimentally derived solutions to technical problems. Twelfth-century authors did not lack habits of observation or an empir­ical attitude; they lacked ideas on how to go beyond empirical results like Philo’s to a theoretical science of nature. They had acute observations, feasible rules of thumb, and were nimble with approximation, but their knowledge remained over-rich in experience and poor in understanding causes. The Alexandrian engineers’ rule of thumb is an example. Could no more rational solution be derived? The problem was not felt. They were glad to have the formula and asked no more. As physicist Steven Weinberg observes, “Nothing about the practice of modern science is obvious to someone who has never seen it done.”⁷