Social Life as Bootstrapped Induction

Abstract

How people refer and how they infer are key empirical questions for the sociology of knowledge. In the present paper, I suggest that in the course of social interaction much referring activity is self-referring, and much inference self-validating. This occurs to the extent that our inductive inferences become permeated with feedback-loops or `bootstraps': I offer a simple general form of representation to assist in thinking about bootstrapped induction. In the second half of the paper I indicate some of the interesting consequences of the existence of bootstrapped induction: I cite the self-fulfilling prophesy as a special case where the induction is destructive, but emphasize the role of bootstrapped induction in constituting stable institutional forms. Finally I raise the question as to how far the bootstraps can be eliminated from patterns of inference: I suggest that this problem might be best attacked by sociologists of natural science.