is hard to argue that America was a social construction and that only by being persuasive and powerful did Columbus and Vespucci convince others of its existence. Similarly, the germ theory worked not because of Pasteur’s political adroitness but because these germs are really there.8 Relativity, quantum mechanics, the structure of DNA are all “discovered,” in this view. The more extreme versions of this positivist and Whiggish approach to the history of science have justly fallen in disrepute. After all, the supporters of phlogiston, the humoral theory of disease, and the ether enveloping the earth were just as certain that they were correct. At the same time, there must be natural regularities and laws we do not even suspect could be true any more than Lord Kelvin suspected quantum theory could exist or than Jean Baptiste Lamarck had any clue of molecular genetics. Somewhere between the nihilism of the social constructivists and the absolutism of the positivists, social scientists and economic historians have to carve out a reasonable position. Constrained contingency seems to me is such a position, and it is perfectly consistent with evolutionary thinking. There was nothing inevitable about the emergence of homo sapiens, of giraffes, or of platypuses. But a lot of life forms are impossible, excluded by the laws of physics, chemistry, or by history. Within the limits of the possible, similarly, useful knowledge and technology emerged. But they picked certain avenues, and not others.
The sources of contingency in the development of technology are thus multiple.9 For one thing, the development of the knowledge underlying technology is itself very hard to predict, as I noted above. Secondly, the institutions governing (1) the emergence of such underlying knowledge, (2) the likelihood of it being applied to new techniques (“invention”), and (3) the chances of such new techniques being “selected” are themselves contingent. We do not have a good theory of why some societies developed one kind of institutions or another, why some adopted the rule of law and constrained taxation, whereas others seem to end up in a very different, low-level state. Clearly, the complex interactive processes that determined what kind of institutions predominate in each society allowed for multiple stable equilibria. The uniqueness of the Industrial Revolution is that it somehow changed the parameters that determined the stability of one subsystem to the point where they violated what economists refer to as the second order conditions. Once that happens, in a global system, no other system – no matter how stable in and of itself – was safe. Tokugawa Japan, central Africa, Romanoff Russia, and Qing China may have been relatively stable societies (at least as far as long-term change was concerned), but once things started to
8For an approach that seems to be uninterested in that fact, see for instance Bruno Latour, The Pasteurization of France. Cambridge, Mass 1988.
9A contingent event is one that requires many necessary causes that are not perfectly correlated with one another. The larger the number of such factors, and the lower their correlation, the more we can define the event as contingent. This is not the same as saying that the event was accidental or unlikely, which requires a statement on the likelihood of the necessary causes to occur.