Pure reason, in both its theoretical and practical forms, has a tendency to run into a certain sort of problem. If one thing depends on another, pure reason expects to be able to trace the dependencies back until it finds the thing that depends on nothing else. However, such an endpoint to any dependency can be found only in the noumenal realm, not in the phenomenal realm. Since the phenomenal realm is the only one we have access to, pure reason is bound to be frustrated.
When pure reason is thus frustrated, it produces "antinomies," conflicting statements both of which appear to be validated by reason. The first Critique contains the antinomies of pure theoretical reason and concludes that to resolve them, we must investigate the workings of pure theoretical reason. Similarly, we will find an antinomy of pure practical reason here that will prove ultimately beneficial since resolving the antinomy will further our knowledge.
The particular series of conditions in question here relate to the good. If an action's goodness depends on something about which it does not itself depend on, what is that? Whatever it is, let us call it the "highest good." To know well enough for practical purposes what the goodness of actions depends on can be termed wisdom. To know—or more modestly, to seek to know—on what the goodness of actions depends in a scientific sense is philosophy, as "philosophy" was understood by the ancient Greeks.
The highest good is the object of pure practical reason. We must carefully draw a distinction, though between the object of pure practical reason and the determining ground we are moved by when we are moved by pure practical reason. The ground of pure practical reason is not the achievement of the highest good. It cannot be, for if it were, one's motivation to follow the moral law would depend on whether or not one cared about the highest good. It is not acceptable for obedience to the moral law to be contingent in that way. Rather, the ground of pure practical reason can only be that of dutifully following pure practical reason.
Kant's uses the term "dialectic" to connote neither a "logical argument" nor a "discussion." His dialectics are arguments that go astray because of some wrong presupposition. Or rather, they are arguments that come in pairs, both of which go astray in opposite ways because of the wrong presupposition. The Dialectic section will then seek to remove the presupposition in order to generate a more justified conclusion about the topic at hand. In this sense, the dialectic is like a discussion, where the two wrong arguments that contain a grain of truth are the two participants. Hegel and Marx model their dialectics after Kant's, where the partial truths of the "thesis" and the "antithesis" are reconciled by the "synthesis" of the two.
If the dialectic of pure practical reason is to run as Kant intends, we must be able to make sense of his distinction between the "object" and the "determining ground" of pure practical reason. That one can create a verbal distinction is clear, but what it might mean is not so clear. Unless we know what Kant means by these terms, we cannot begin to say whether he is right about what object and what determining ground are appropriate to pure practical reason.
The distinction might be that the object is the motive of the act, either in the sense of the ultimate goal of the act or in the sense of what the agent has in mind while performing the act. On the other hand, the determining ground could be what determines whether or not we perform the act. If we consider a person who rescues a drowning baby, Kant might be saying that the object is the highest good in the sense that the object is what the person is considering while they perform the rescue, or in the sense that the person's ultimate goal of acting selflessly is the highest good. Dutifulness, on the other hand, is what determines whether or not the baby will be saved, and also whether or not these other mental attitudes will be held.
It is also possible, though, that Kant would say rather that the highest good is the object, in the sense of being the conscious aim, while dutifulness is the determining ground, in the sense of being the ultimate goal. It is also possible that none of these distinctions is quite what Kant has in mind.
In order to clarify the notion of an antinomy, let us consider one of the antinomies from the first Critique. Events in the world are always caused by other events in the world. The antinomy of freedom asks whether there is a first cause. If there is, this is a problem, for it is itself uncaused, so there must not be one. If there is not one, then we must comprehend an infinite series of causes as having already taken place, and since we cannot do that, there must be a first cause after all. The solution is to distinguish the noumenal and the phenomenal. The first cause exists, but only in the noumenal realm and there is no problem with noumenal uncaused causes. And yet it is not true that there is an infinite sequence of causes, for the phenomenal extends only so far as one happens to have had experiences of it. Since we grasp a finite but indeterminately extendable amount, there is no question of an infinite uncaused series.
Now we can anticipate what is to come. Good actions depend on the highest good to make them worthwhile. Assuming that there is a highest good leads to paradox, as does assuming that there is no highest good. The solution lies in reference to the noumenal world.