of the process, but by systematic trial and error settling on techniques that work.26 These methods are never as efficient as those based on sound and deep principles, but they themselves can be made to work better through the methods of parameter variation and experimentation models, and today through computer simulation. This methodology has been attributed to the great eighteenth century engineer, John Smeaton, whose impact on the Industrial Revolution was no less than Watt’s despite his lesser fame.27 A century later, William Rankine formalized engineering science in these terms, and made it a respectable discipline at the University of Glasgow.28
The framework I delineated above, and through which western technology expanded beyond any conceivable boundaries differs from the standard accounts that attribute its success to the scientific breakthroughs of the nineteenth century. Very few techniques are ever designed from first principles alone. Epistemic bases are almost always too narrow to work this way, even relatively late in the development stage. Instead, there is always a great deal of experimentation and trial-and-error that reflects the inability of “pure theory” to take into account all the variables that come into play in “real” situations. The point, however, is that a wider base in useful knowledge allows people to rule out larger and larger areas that are excluded, restrict the searches and thus make them more efficient. Such knowledge prepare the minds that fortune (and hard work) favors. An evolutionary framework suggests that if nothing is known, the search will be totally random, much like happens in a pure Darwinian framework. In the other limit, if everything is known, we need not experiment, since we can design what we need optimally. Historically, all societies find themselves between these extremes, but the Industrial Revolution and the unprecedented growth of technology rested on moving on the line in between, perhaps from a world in which very little was known (before) to one in which a little more was known. That may have made all the difference.
Between 1750 and 1850, all of these three negative feedback mechanisms were gradually transformed and neutralized or turned into positive feedback. We may call this the Industrial Revolution, but the sources of these changes go back to the institutional changes we associate with the enlightenment. We do not usually associate the enlightenment and the scientific revolution that
26Nathan Rosenberg, “Engineering Knowledge,” unpublished ms., Stanford University, 2001.
27Smeaton is credited with the development of the method of parameter variation through experimentation, which is a systematic way of gradual improvements in the absence of a wide epistemic base. It establishes regularities in the relationships between relevant variables and then extrapolates outside the known relations to establish optimal performance. Vincenti (1990, pp. 138-140) and Cardwell (1994, p. 195)
28David F. Channell, 1982. “The Harmony of Theory and practice: the Engineering Science of W.J.M. Rankine.” Technology and Culture Vol. 23, No. 1 (January), pp. 39-52.