"Readers' Forum, Comments on 'The Tradition of Spontaneous Order' by Norman Barry"

2. Mario J. Rizzo,[*] “Spontaneous Orders: Deterministic or Nondeterministic?”

[I]f there is nothing unforeseen, no invention or creation in the universe,
time is useless…. For time is here deprived of efficacy, and if it does
nothing, it is nothing.

—Henri Bergson[1]

There are two forms of spontaneous order theories which I wish to distinguish in this brief note: those that relate to the origin of an aggregate structure and those that involve the function of the structure.[2] The common element present in all theories of the first type is the claim that some overall social patterns or institutions are caused by a myriad of decentralized actions that do not aim at their establishment. Theories of the second type, however, disregard the origin of the pattern and seek, instead, to explain why it continues in existence. These functional theories recompose the structure in terms of the purposes it serves for the individual. Presumably, these will explain why the individual actions that give rise to the aggregate structure will themselves endure and hence why their product endures.

The claim I shall make is simply this: theories of spontaneous order, whether of the first (origin) or second (function) variety, cannot be deterministic if they are to explain economic or social processes over time.

Suppose, for example, we were to adopt the position that the causal link between decentralized actions and social structures or orders is deterministic. Then, on this assumption, certain initial conditions (actions 1…n) in conjunction with a theoretical law would yield with logical necessity the structure we want to explain. This rigid link between initial conditions and result is radical mechanism.[3] Such explanations cannot tell the story of how orders can arise in the course of time. Instead, they can only provide a logical or static recomposition of an already-arisen order. For if the connection between cause and effect is deterministic then time literally adds nothing. Thus the aggregate structure should have already existed from day one but it did not. By the principle of causality, then, time must add something. This something is the future decisions and choices of the many acting individuals. Since these decisions cannot be predicted by those who will make them,[4] we cannot model the individuals as foreseeing the emergent order. Hence genuine uncertainty or “surprise” must be part of any methodological individualistic story of the origin of social institutions.

Spontaneous order theories of the functionalist variety sometimes claim that the function which an institution serves provides a logically sufficient explanation of why it continues to exist. This claim is just the inversion of radical mechanism or, simply, radical finalism.[5] Instead of temporally antecedent events rigidly determining current institutions, we postulate that future functions determine (or explain) them. Since individuals act on the basis of their anticipations, it is only the future (anticipated) functions of institutions that could possibly be relevant. Such functionalist theories cannot, however, be evolutionary in the true sense.[6] This is because the complete set of sufficient conditions that maintain an order are created in the evolutionary process itself.[7] Time must add something. In this case, what it adds is a change in individual knowledge and the anticipation of a possibly better way of achieving one’s purposes. Thus, “order [is] defined in the process of its emergence.”[8] In retrospect, when the complete set of causes is known (at least in principle) we might find it useful to construct a model of evolutionary process as aiming at some determinate function. Nevertheless, this model is only a heuristic delusion and may well lead us astray if we are not extremely careful. Ex ante (in advance), any truly evolutionary process is itself a part of the ultimate outcome.

The general conclusion that can be drawn from these arguments is that theories of spontaneous order (and, a fortiori, of equilibrium) must be pattern explanations.[9] The conjunction of statements about initial actions and a law explains the overall pattern or class of existing institutions rather than any specific institution. Similarly, functional theories can rationalize the class of possible structures that will serve a particular function rather than ‘postdict’ the optimal structure. As John von Neumann and Oskar Morgenstern have said, “[T]he complete answer to any specific problem consists not in finding a solution, but in determining the set of all solutions.”[10]