document that tried to point out this basic point. It seems reasonable to surmise that much of the economic decline of the Iberia in the sixteenth century and that of the United Provinces in the eighteenth can be explained in part by such mechanisms. Perhaps the most insidious of these forms of institutional feedback was organized resistance to new technology by vested interest groups, a phenomenon still quite visible in our own time.25
The third mechanism is the one most relevant to the interpretation I am suggesting here. Simply put, before 1750 the vast bulk of techniques in use anywhere in the world rested on very narrow epistemic bases that constituted a “fixed” factor that lent the system a kind of concavity that we do normally not associated with knowledge. But what people used in production was engineering without thermodynamics or mechanics, iron-making without metallurgy, farming without organic chemistry, and medical practice without physiology or microbiology. What was known, had been discovered serendipitously or through trial-and-error experimentation. The point is not that the operators of techniques (that is producers) themselves were unaware of the principles of physics and chemistry that underlay the techniques they carried out; that remained true much later and is largely the case in our own time as well. The point is that nobody knew. Techniques that worked, from animal breeding to steelmaking, soon reached ceilings that might have been broken through had someone understood a bit more as to why they worked. The processing and manipulation of materials, the use and design of instruments and machines, and the raising of edible crops and animals, were little informed by theory, because there was little theory.
Needless to say, the marginal product of widening these epistemic bases varied from activity to activity. A dexterous and brilliant mechanic like Watt or Roberts might make very substantial improvements in a piece of equipment without a complete understanding of why this turned out to be correct. But in organic chemistry, the use of electromagnets in telegraphy, or the design of steam turbines, there simply was no substitute for formal science. Moreover, the Industrial enlightenment created ways of getting around narrow epistemic bases. It became gradually clear that it was possible to generate enough data to catalog and describe patterns that some phenomena could be harnessed and exploited even without a thorough understanding. What Layton has termed “engineering science,” seeks a more concrete level of knowledge of how things work in actual situations, under conditions of varying pressure, stress, heat, and friction rarely examined by pure theorists. This is what Nathan Rosenberg calls “engineering knowledge” – based on a limited and perhaps even shallow understanding
25See Joel Mokyr, “The Political Economy of Technological Change: Resistance and Innovation in Economic history” in Maxine Berg and Kristin Bruland, eds., Technological Revolutions in Europe, 1998 , pp. 39-64; ibid., Gifts of Athena, ch. 6.