I really appreciated what Dr. Vern Poythress wrote in his book “Science and Hermeneutics” on Immanuel Kant and his thoughts relating to science.
This third way is in many respects the most promising. As we observed, the procedure of using specific scientific theories is useful only when a specific theory happens to touch on issues of human concern. Most of the time it does not. The procedure of building a world view is questionable, since one must extrapolate science beyond what has been verified. On the other hand, the procedure of building an epistemology relies on the undoubted success of science as a means for producing knowledge. Even if science does not include all knowledge, its success surely contains lessons that apply to all knowledge.
The classic example of using science as a platform for epistemology is to be found in Immanuel Kant.10 Kant at an early point in his life followed the rationalistic, deductivistic approach of Leibniz. In opposition to this rationalism Hume defended an empiricism that started with pure events and did not assume that they were connected, merely that they sometimes occurred together. As Kant testified, Hume “awakened him from his dogmatic slumber.” Kant then rejected rationalism. He was convinced that there was no guarantee that phenomena would turn out to be connected in the way that a rationalist supposed. And yet Hume’s own empiricist solution was also inadequate. Hume’s world contained no intrinsic connection between individual events. Hence Hume could not account for the reliability of scientific knowledge. The rationalistic approach of Leibniz did not lead to fruitful science either.
Kant therefore endeavored to provide an epistemology that was adequate to science and that also preserved room for religion. Kant accepted the obvious fact that science did provide knowledge. Kant’s task was then to provide an epistemology that accounted for the success of science. Science arose from a combination of empirical data (Hume’s concern) and rational inference (Leibniz’s concern). An adequate epistemology would do justice to both these elements.
Kant’s solution was to say that, whatever we observed empirically, we observed necessarily in terms of categories presupposed by the human mind. Inner experience (experience even within one’s mind, without looking at the world) was necessarily experience against the background of time. Outer experience was necessarily experience in a framework of both time and space. To these categories of time and space one could also add the categories of quantity and causality, which are basic to physics. The empirical element in science was accounted for, since human experience was experience of a world outside that was not always predictable. On the other hand, the rational element in science was accounted for, since human experience necessarily conformed to the preestablished categories of the mind. The world of phenomena was not pure confusion, as Hume had it. Rather, it was necessarily a world of time, space, and causality, and this was the foundation for sure knowledge.
On the surface, Kant’s solution seems attractive. In fact, however, a closer examination shows that it provides both too little and too much for the needs of physical science.11 On the one hand, it provides too little. Suppose we agree that it shows the necessity of conceiving the world in terms of the categories of time, space, causality, and quantity. This result still does not constitute scientific knowledge, nor is it an adequate basis for guaranteeing that we can obtain scientific knowledge.
After all, any particular physical theory, such as Newton’s laws, Boyle’s law, or Dalton’s law, furnishes predictions to the effect that the world will behave in one way, not another way, within the general framework of time, space, and causality. To say that there are causal connections (Kant) is not yet to say what kind of causal connections there are (Newton). To say that everything has a cause (Kant) is a long way from saying that all bodies attract one another with a force given by the formula F=GmM/r\S2\s (Newton). Kant’s epistemology guarantees only that there is necessarily a cause for any event. It does not allow anyone to say that the cause must necessarily be what Newton says it is. In fact, in Kant’s scheme the particular way that the world is, within the conceptual framework of causality, is ultimately not predictable. It is contingent. The phenomena presuppose things in themselves that cannot be predicted beforehand. A tightly formulated general law, like Newton’s, predicts something that Kant says cannot be predicted. Hence Kant still cannot explain why a simple formula like F=GmM/r\S2\s should hold true all the time, while other formulas do not.
Second, Kant’s epistemology provides too much for the needs of science. Namely, it dictates to science assumptions that may not turn out to be factually correct. Kant’s categories of time, space, and causality are most naturally understood as actually implying a particular theory of physical time, space, and causality. In Kant’s environment, these categories seemed to imply a linear absolute time scale, Euclidean space, and determinism in the realm of physical causes. These things were virtually part of people’s intuitions about space, time, and causality. These intuitions or views of time, space, and causality were compatible with Newton’s theory of gravitation, and so people were content with them at the time.
In the light of better scientific knowledge, however, physicists today are not willing to agree with Kant. Physicists now realize that the ideas of absolute time, Euclidean space, and determinism were all assumptions about the world that might be either true or false, not presuppositions that were necessarily true. Kant did not take into account two facts. First, the psychic experience of time, space, and causality by the ordinary person is not the same as the time, space, and causality that may be most suitable to physical theory. Intuitions derived from psychic experiences may or may not be immediately useful in physical theory. Second, intuitions themselves can be reformed. In Kant’s time “space” meant Euclidean space. Given a line and a point not on the line, one and only one line could be drawn through the point, parallel to the first line. But modern physicists, confronted with coherent alternatives to this scheme, have had their intuitions changed. For them it is not obvious (in fact, it is false) that physical space is Euclidean.
Kant’s solution, then, did not really correspond well with the nature of science. It did not even fit the specifics of Newtonian science. And it fitted even less well the developments of the twentieth century that were destined to supersede Newton. But Kant’s philosophy had enormous impact nonetheless. It was accepted because of its promise in the field of philosophy rather than because of its accuracy in the realm of science.
For Kant, in fact, epistemology became the basis for philosophy as a whole. By means of his reflection on the categories of the human mind Kant specified what could and could not be the object of knowledge. And this pronouncement virtually determined what could and could not be part of the world. From Kant until the twentieth century, epistemology has been the key to philosophy as a whole. Hence Kant’s work was not just an epistemology. It was a full-blown philosophy. It provided its own world view.