as epistemic bases.13 But the form that the search for this knowledge takes, the effectiveness and results of the search, and the way societies use this knowledge are all quite variable.
In any event, asserting that certain natural regularities are knowable does not guarantee that they will be known. For that to happen, three conditions have to hold. First, society needs to create the incentives that make it worthwhile to investigate nature. Individuals who discover natural facts or laws that are in some way useful need to be rewarded by society or at least not punished as potential heretics or troublemakers. Such rewards maybe purely financial through the patronage of wealthy and powerful sponsors, as was often the case during the seventeenth and eighteenth centuries, or some sinecure like a tenured professorship or a pension. Societies that rewarded distinguished scientists, promising a great deal of social prestige by elevating them to a peerage or awarding them Nobel prizes for breakthroughs in useful knowledge will find themselves allocating more resources to natural investigations and have a better chance at expanding their epistemic bases. Yet the intellectual property rights to this knowledge need to be arranged with caution, since excluding others will violate welfare maximization given the non-rivalrousness of knowledge, and will reduce sharply the chances of this knowledge being applied.
Second, the research agenda needs to reflect topics that can serve as the epistemic base for growth-inducing inventions. These agendas are of course to some extent set by social priorities. Researchers realize that progress in areas of high social demand, such as cancer research or Alzheimer’s disease will be more rewarded than the reproductive habits of amphibians. Such priorities need to be signaled clearly through the incentive structure and they create what may be called “induced research.” But they alone do not set the agenda. What matters too is that techniques in use themselves raise issues of why and how they work. They serve as focusing devices, to use Rosenberg’s term. History offers many examples of such research being inspired by techniques with a narrow epistemic base, widening the base, improving the technique and so on. This kind of cumulative, divergent positive feedback mechanism between the epistemic base of a technique and the technique itself is widely observed in post-Industrial Revolution history, but rare before. I already mentioned the telegraph. Another classic example is the use of aspirin, for many decades a technique resting on a very narrow epistemic base indeed. Only in the 1960s did John R. Vane, Bengt Samuelsson, and Sune Bergström discover the operation of prostaglandins, and with it the development of a whole new series of analgesics that
13I am abstracting here from the problem who “we” are. Suffice to say that the formal definition of social knowledge os the union of all sets of individual knowledge, so in principle if only one individual in society knows a particular fact or regularity, he or she can transmit this to others who can then formulate techniques based on it.