One rational standard of action is how well it promotes the end it seeks. Another standard is whether it aims at ends which are good. Both of these, but especially the former, depend on judgments of fact.

My own view is that Kant's conception of the duality of the good (morality and happiness, the good of our person and the good of our state or condition) is a distinctively modern view.

As I understand it, Kantian constructivism is partly a position in normative ethics and partly a position in metaethics. In metaethics, it is the position that ethical claims have truth values, but their truth conditions consist not in a set of objective facts to which they correspond, but instead in the outcome of some procedure of deliberation resulting in decisions about what to do.

Kant regards the universalizability test for maxims as focused on a very special sort of situation: one where the agent is tempted to make an exception to a recognized duty out of self-preference. The universalizability test is supposed help the agent to see, in a particular case of moral judgment, that self-preference is not a satisfactory reason for exempting yourself from a duty you recognize. Kant thinks, as a matter of human nature, that this situation arises often enough and that we need a canon of judgment to guard against it.

Not only our moral life, but even our use of theoretical reason - on which we rely in rationally inquiring into nature - presupposes that we are free.

There is a tradition of "modernist" theology arising out of post-Kantian thought - Fichte was the real father of it, but Schleiermacher and others also developed it - which might have more promise if it had greater influence on popular religion.

In fact people do not know enough about themselves and what is good for them to form a sufficiently definite conception of the general happiness (or whatever the end is) to establish definite rules for its pursuit.