Abstract

The major concern of this thesis is to provide a plausible argument for the existence of another goal of science--so that van Fraassen's claim that empirical adequacy is the sole aim of science can be countered. ;I propose that we should look carefully into why theories are constructed in order to find some clue to the problem of how another goal of science can be established. More specifically, we should examine closely what unification in theories is, in order to see the limitation of treating science as an activity aiming only at empirical adequacy. It is suggested in this thesis that unification is best understood when the way hypotheses in a theory relate to one another is made explicit; and that confirmation pathways are what reflect this relation. Yet, if van Fraassen is right in claiming that all that really matters in adjudicating theories is empirical adequacy, having complicated confirmation pathways between hypotheses and evidence is by no means a desirable feature one should look for. In other words, under the view that empirical adequacy is the only aim of science, so long as two theories are empirically equivalent, it does not matter how the hypotheses in each are connected to provide confirmation pathways. On the other hand, we find that the Glymourian concept of confirmation can provide a good focal point for demonstrating why empirical adequacy is not all that science takes as its aim. Glymour's concept of confirmation takes into account of the degree of complication of confirmation networks in theories and, as a result, it can be utilized to demonstrate the desirability of having more confirmation pathways running in all kinds of ways between hypotheses and evidence. Ultimately, that leads to the conclusion that empirically equivalent theories are not necessarily equally desirable, and greater confirmability stands for another criterion for adjudicating theories. That is, we have higher degree of confirmability as another goal of science. ;Some implications of the present perspective concerning issues such as lawlikeness, realism, explanation and scientific acceptance will be drawn, and they will be contrasted with the ways these issues are discussed in van Fraassen's theory