Instantiation, Confirmation, and Truth: A Problem in Inductive Logic
Dissertation, The University of North Carolina at Chapel Hill (
1980)
Copy
BIBTEX
Abstract
Many writers have felt that the inclusion of such evidential factors into confirmation theory must sever any connection between science and truth. Others, such as Quine, have offered faulty arguments to establish such a connection. It is seen that the inclusion of these noninstantial forms of evidence does not sever science from truth, given a proper understanding of these evidential considerations and a realistic perspective on science. ;Confirmation theories based solely on instantiation are inadequate for the modeling of science. Nevertheless, it is worthwhile to see clearly why they fail. Qualitative theories are not adequate for the comparison of hypotheses, nor for basically comparative forms of evidence, such as simplicity. Quantitative instantial theories face an infinite regress when it comes to evaluating the necessary parameters. ;Given the inadequacy of instantial confirmation theories, noninstantial forms of evidence are examined. Goodman's theory of projectibility is found to be inadequate, although it is motivated by a justifiable principle of conservativism. Ultimately, it is seen that all noninstantial evidence can be reduced to three basic types: generality, simplicity, and information content.