Abstract
FOR METHOD; POSSIBLE EPISTEMOLOGICAL STANDARDS IN A POST-POSITIVIST ERA.
In this essay I take seriously the logical problem about the inference of theories from the observations that we make in the world - the thesis of underdetermination. Given the truth of this thesis we must question both the order in the world and the order that acts as a premise for our understanding and knowledge of that world. This is a general problem for all the sciences, but in this essay I shall concentrate upon this problem as it emerges in parts of the social sciences.
If the thesis of underdetermination makes sense to us, one consequence is that we may not presuppose that the processes in this world are fully determined. We will have to allow the possibility that some of the processes are indetermined. We must allow that some processes of this world are open, nonstable, out of equilibrium, and constantly changing. In such a world regularities will have the character of 'if p, sometirnes q'. For such processes not even Laplace's demon will be able to know what will happen.
Such a possible metaphysics will have consequences for our epistemological and methodological order. The machine-paradigm with its determinism views our ignorance about the world as the consequence of human inability. If some of the processes are seen as open and indetermined, indeed a reasonable picture for the biological and social sciences, we need to adjust our epistemology and methodology we will have to adjust our expectations about what constitutes adequate scientific explanations.
Given that the world contains indetermined processes, the epistemological and methodological order must be able to absorb such a possibility by offering theories of explanation that allows genuinely nondeterministic regularities. The most commonly known theory setting the standards for scientific explanation the theory of Carl Hempel is basically for a deterministic world, and, although an honorific and highly idealistic theory, it cannot, due to its deterministic assumptions and logical structure, judge explanations of indetermined processes as adequate. We need a theory that both allows indetermined processes as a standard possibility of the world and that is tied closer to actual explanatory processes. I shall in this essay exhibit the theory of Wesley Salmon which may be said to do the job. I shall also show how we may change the focus of our demands for scientific adequacy by showing how post-empiricist standards of explanatory adequacy may work well in a social science that allows for genuinely indetermined processes.
As a vehicle for this demonstration of epistemological standards I shall use the social science problem of uncoerced social and political order as it is presented and accounted for in two texts: Jon Elster's The Cement of Society (1989) and Jarnes March' and Johan P. Olsen's Rediscovering Institutions (1989). Perceiving uncoerced order as just such an indetermined process, with regularities on the form 'if p, sometimes q', the explanatory strategies in these texts will be submitted to both Wesley Salmon's theory of explanation and to one post-empiricist standard of explanatory adequacy. This is done in order to show how we may both continue to expect and to claim an epistemology and a method for the social science problem of uncoerced social and political order.