Kant's most original contribution to philosophy is his "Copernican Revolution," that, as he puts it, it is the representation that makes the object possible rather than the object that makes the representation possible. This introduced the human mind as an active originator of experience rather than just a passive recipient of perception. Something like this now seems obvious: the mind could be a tabula rasa, a "blank tablet," no more than a bathtub full of silicon chips could be a digital computer. Perceptual input must be processed, i.e. recognized, or it would just be noise -- "less even than a *****" or "nothing to us," as Kant alternatively puts it.
But if the mind actively generates perception, this raises the question whether the result has anything to do with the world, or if so, how much. The answer to the question, unusual, ambiguous, or confusing as it would be, made for endless trouble both in Kant's thought and for a posterity trying to figure him out. To the extent that knowledge depends on the structure of the mind and not on the world, knowledge would have no connection to the world and is not even true representation, just a solipsistic or intersubjective fantasy. Kantianism seems threatened with "psychologism," the doctrine that what we know is our own psychology, not external things. Kant did say, consistent with psychologism, that basically we don't know about "things-in-themselves," objects as they exist apart from perception. But at the same time Kant thought he was vindicating both a scientific realism, where science really knows the world, and a moral realism, where there is objective moral obligation, for both of which a connection to external or objective existence is essential. And there were also terribly important features of things-in-themselves that we do have some notion about and that are of fundamental importance to human life, not just morality but what he called the three "Ideas" of reason: God, freedom, and immortality. Kant always believed that the rational structure of the mind reflected the rational structure of the world, even of things-in-themselves -- that the "operating system" of the processor, by modern analogy, matched the operating system of reality. But Kant had no real argument for this -- the "Ideas" of reason just become "postulates" of morality -- and his system leaves it as something unprovable. The paradoxes of Kant's efforts to reconcile his conflicting approaches and requirements made it very difficult for most later philosophers to take the overall system seriously.
Nevertheless, Kant's theory does all sorts of things that seem appropriate for a non-reductionistic philosophical system and that later philosophy has had trouble doing at all. Kant managed to provide, in phenomenal reality (phaenomena="appearances"), for a sphere for science that was distinct and separate from anything that would relate to morality or religion. The endless confusion and conflict that still results from people trying to figure out whether or how science and religion should fit together is deftly avoided by Kant, who can say, for instance, that God and divine creation cannot be part of any truly scientific theory because both involve "unconditioned" realities, while science can only deal with conditioned realities. In the world, everything affects everything else, but the traditional view, found even in Spinoza, is that God is free of any external causal influences. Similarly, Kant can be a phenomenal determinist with science yet simultaneously allow for free will, and that in a way that will not be entirely explicable to us -- a virtue when the very idea of a rational and purposive free will, and not just arbitrary choices, has involved obscurities that no one has been able to resolve. Kant's theory prevents psychological explanations for behavior, however illuminating, being used to excuse moral responsibility and accountability. Thus, the tragic ********* of the defendant, however touching and understandable, cannot excuse crimes commited in full knowledge of their significance.
