1. 1. According to Maine: 'It is more likely that joint ownership, and not separate ownership, is the really archaic institution, and that the forms of property which will afford us instruction will be those which arc associated with the rights of families and of groups of kindred' [Maine, 1877 , p.259 ].
2. 2. In his study of Roman law, Noyes[1936 , pp.94 -97 ] offers the observation that the first property, economically and legally, was property of the family; concepts of communal property developed later.
3. 3. Of course many of the conceptual elements used in the paper have been foreshadowed in the work of others. Reference to Dahlman's[1980] work on English common fields is especially appropriate. He does not, however, try to develop an integrated model out of these elements, perhaps because conceptually he uses an ideal-type approach; that is, he concentrates on forms which in the present model are just two points on a continuum. The same is true of Umbeck's[1981] treatment of property rights in the California gold rush.
4. 4. Even though rules are discrete as written, however, the level of enforcement is a continuous variable, implying that most de facto institutions are better approached with continuous techniques.
5. 5. These assumptions appear to put us into the realm of 'non-exclusive private goods', in the language of Blumelet al. [1986]. But goods of this type always involve pure public goods, defined as those involving joint consumability. The item being jointly consumed by users of the resource is restraint; the benefits produced by the restraint of individual resource users are conferred jointly on all users. As in any public good, the total amount of restraint supplied is likely to be inefficient on the low side.