Abstract

The title of my dissertation is "confirmation theory & confirmation logic", and it consists of five Parts. The motivation of the dissertation was to construct an adequate confirmation theory that could solve "the paradoxes of confirmation" discovered by Carl G. Hempel. ;I try mainly to do the three things: introduce the fundamentals of Hempel's theory of qualitative confirmation as the common background for subsequent discussions, review the major views of the paradoxes of confirmation, present a new view, which is more radical than other known views, and argue that a solution to the paradoxes of confirmation may require a change of logic. ;In Part Two I construct a number of promising three-valued logics. I employ these "quasi confirmation logics" as the underlying logics of some new confirmation theories which, I had hoped, would solve the paradoxes of confirmation. I consider three-valued logics instead of any other many-valued logics as the underlying logic for any promising confirmation theory, because I believe that there is some intimate relationship or, even, a one-to-one correspondence between the three truth-values of "truth", "falsity" and "neither truth nor falsity" and, respectively, the three confirmation-statuses of "confirmation", "disconfirmation" and "neutrality". ;In the last Part I try to appraise the three most important confirmation theories discussed and/or constructed in this dissertation. They are Hempel's theory of confirmation, Goodman's and Scheffler's theory of selective confirmation and the internal confirmation theory. ;After some more vigorous criticisms are made and some new paradoxes of confirmation are unexpectedly derived in both the theory of selective confirmation and the internal confirmation theory, I arrive at, perhaps reluctantly, this more reasonable conclusion: under the present situation when there is no obvious way to overcome the new difficulties the best thing that we can do is to dissolve all--new and old--paradoxes of confirmation, for Hempel may be after all right to say that the paradoxes of confirmation are not genuine and to think otherwise is to have psychological illusions as Hempel says