are contracts?' and so on, and it was often very hard to know what in practice the concepts that were being employed really meant. ( Coase 1992: 333)
Coase goes on to call for more empirical work, including the collection of contracts, 'so that in future we do not have to pretend to know their characteristics and therefore devote ourselves to studying imaginary contracts rather than real ones'. But Coase is certainly not calling for some simple empiricism; he also makes a methodological point about theoretical modelling which, although he doesn't say so himself, would undermine much of the work on economic theory done in this area to date:
I also want to say something else about modelling . . . I think you can be too precise too soon and that this is a situation you are liable to be in when you are very ignorant. And I think that we are very ignorant in this field . . . To have a model that simply incorporates what you know (or think you know) at an early stage may, in fact, by producing results that are very misleading, prevent useful research from taking place . . . I fear that we might be in this sort of condition now where people will produce models in which we ignore what subsequently are discovered to be important aspects of the problem, but which, in the meantime, put us off doing the research necessary to find them. ( Coase 1992: 335-6) 9
Examples of such important aspects might include the concepts of 'knowledge', 'trust', and 'power'. These are surely central to an understanding of contracts and their use, the firm and its operation, and industrial organization more generally, when firms go out into the 'market', but rarely as passive price-takers. Instead we see co-operation, collusion, and various other relationships, including the development of joint ventures, networks, and clans.