Kant thinks of laws of nature as descriptive, universal and deterministic; e.g., Boyle's Law. The phrase "Laws of freedom" sounds self-contradictory. But for Kant, freedom entails neither license nor arbitrary choice, so it, too, will exhibit lawfulness. Although universal, any laws of freedom will turn out to be radically unlike laws of nature by being prescriptive and non-deterministic. How such laws are possible poses, for Kant, deep and difficult problems. These problems aren't addressed until the end of Section II and Section III. The problem of freedom is also discussed in Kant's Critique of Pure Reason and his Critique of Practical Reason
Logic consists of purely formal laws; e.g., if p then q, not-q, so not-p (modus tollens). These laws have a necessity that cannot be provided by experience. The natural sciences and ethics, however, are not entirely formal. While the principle that "every event has a cause" is synthetic yet known a priori --and hence is not based on experience--the specific content of the natural sciences is derived from experience; e.g., that smoking causes lung cancer. The empirical part of ethics includes what we now call psychology and the social sciences. These are important because they tell us what kinds of challenges we face as we think and judge what we are to do. Whatever laws either the natural or the social sciences confirm, however, has only a limited bearing on morality. Clearly, if you had been pushed off the top of the Empire State Building, you will "obey" the law of freely falling bodies. As you cannot do otherwise, it would be fatuous to tell you that you "ought" to fall faster or slower or not at all. But Kant does not believe that psychological or sociological laws eliminate freedom and responsibility. So a law stating that everyone breaks promises from time to time is, strictly speaking, irrelevant to any deliberation about what I ought to do or whether I'm culpable for any lies I tell.
In his Critique of Pure Reason Kant argues that speculative metaphysics makes extravagant claims leading to antinomies; i.e., we can demonstrate both a thesis and its denial by impeccable arguments. Metaphysics, however, need not be speculative: it can be critical. It can, that is, seek to discover those Intuitions (of space and time) and Categories (e.g., substance, causality) without which we cannot make sense of experience. Consider our judgment that every event has a cause. Because it is not merely definitional, it is not analytic; it is synthetic in that it connects two things separable in thought. As it is a synthetic we cannot demonstrate its truth by pure logic, as speculative metaphysicians try to do. Nor, as David Hume shows, can we derive that judgment from experience. Hume's conclusion raises the spectre of skepticism. Fortunately, we can acquire justified a priori knowledge of this judgment because of its role in making experience possible for us: the judgment is both universal and strictly necessary. Establishing the validity of such principles is part of pure philosophy, though not pure formal philosophy (which is formal logic as we know it).
That part of science that considers and establishes certain pure, categorical principles of experience Kant calls a metaphysics of nature. As we've seen, such a metaphysics will be critical, not speculative. The aim of the Grundlegung is to determine whether a metaphysics of morals is possible. If it is, then we should be able to uncover one or more fundamental laws of freedom that determines how we are to act in a way analogous (but not identical) to the way the judgment that every event has a cause determines how we order the world as experienced.
Traditionally, metaphysicians have tried to demonstrate the existence of God, freedom, immorality and to prove that general terms -- e.g., beauty -- did or did not refer to some transcendent reality. None of this, Kant argues, can be demonstrated. There is, however, an a priori element to a metaphysics of nature which can be made out; viz., forms of understanding (e.g., causality, substance) and forms of judgment (e.g., that every event has a cause, that every property belongs to some substance) which must be presupposed if we are to have experience we undoubtedly have.
Just as a critical, or transcendental, metaphysics of nature shows how experience is possible, so a metaphysics of morals attempts an a priori grounding of the ultimate foundation for morality. It will succeed if and only if reason can establish a fundamental principle governing conduct for any rational being. This task is, if anything, more onerous than that faced in the Critique of Pure Reason. For a critical metaphysics of morals, Kant recognizes, depends on the reality of freedom in a way that a critical metaphysics of nature does not. Kant considers this problem in Section 3 and the last bit of Section 2.
In this paragraph Kant notes that we all do, in fact, have a common idea of duty and of moral laws. As is typical, Kant asks what this common idea presupposes. Kant's answer is that we all agree that any moral law must bind, and bind absolutely, if it is to serve as the basis for obligation. And if true, then any moral law must have a necessity that cannot be derived from nature or experience. First, experience -- even experience that is never disconfirmed -- never amounts anything more than a posteriori knowledge. It lacks what Kant calls "strict necessity." There can be precepts based on experience -- e.g., "honesty is the best policy" or "a penny earned is a penny saved" -- but these don't obligate us: why must I save my pennies if I can make a killing through gambling? Second, Kant thinks that a moral law, if one is possible, holds for all rational beings (if there are others than us). In all likelihood, such beings would have a different "anthropology" than do we. Since our obligations cannot be derived from nature or experience, they must be grounded in rationality.
Throughout the Grundlegung Kant insists on the importance of this distinction. One who conforms to any law (moral or legal) does the right thing, but unless one conforms to the law because its the law, Kant worries -- not without cause -- that the conformity will be "only very contingent and uncertain." For if one conforms only because of praise, blame, threats, punishment, self-satisfaction and the like, then if stronger, contrary motives arise, one may feel justified in setting the law to one side. For why conform to any law if it isn't in one's interest?