It is the fourth section of Groundwork III that is headlined "how is a categorical imperative possible?" and it is here that the focus on the reading of the "deduction" of the categorical imperative is based. As Allison puts it: "Since every imperative involves a necessitation...of the will, to account for its possibility is to account for this necessitation" (332). With regard to the categorical imperative we have a synthetic a priori proposition and accounting for the necessitation requires showing that the imperative is, as Allison puts it, "practically possible" by which is meant that it expresses something whose binding character can be shown.
In the opening paragraph of the fourth section of Groundwork III Kant discusses the sense in which rational beings are forms of "intelligence" and indicates that if we were merely members of the world of understanding all our actions would conform 'perfectly' with the principle of autonomy of the pure will. Kant also goes on to claim later that the world of understanding contains the "ground" of both the world of sense and of its laws and purports to use this claim to show that as a being of intelligence I am therefore bound by the laws of the world of understanding.
Allison's analysis of the "deduction" argument begins with the claim that thinking beings place themselves in the world of the understanding but adding to this the point that it is only if such beings also have will that the argument can begin. The next step of the argument, for Allison, is the claim that the self, considered as an active being, possesses spontaneous causal power so that, when viewed as an intelligence, it is seen to be a causal agent. The third key step is that actions, when considered thus to be originated by the agent in question so viewed are part of the "space of reasons". Fourthly, this aspect of the self is contrasted with the empirical one which is governed by reference to "happiness".
Allison's major interpretative move, however, concerns the fifth and final step of the argument as he views it, a step he breaks down into two sub-steps. The first sub-step is concerned with the claim that the world of understanding contains the "ground" of that of sense. The key question with regard to this claim concerns the way the grounding relation is to be understood. Following his general deflationary manner of reading Allison dismisses the idea that this should be seen in a "metaphysical" way as a causal claim. However in addition to the general philosophical reasons Allison possesses for resisting this reading he adds an important normative one to the effect that such a reading provides no binding ground for accepting the validity of the "laws" of understanding. What is needed instead of such a reading is thus one that does show the ground of validity of these laws.
Hence it is not only "ground" that needs to be given an account of but also "laws" in the sense that there could be a "law" of understanding that was validly taken to be binding by a sensible being. In the Critique of Pure Reason this question drove the argument of the transcendental deduction, the schematism and the general Analytic of Principles. From the argument of this part of the Critique we arrived at an account of the conditions of possibility of experience and the analogy between this procedure and that undertaken here in Groundwork III would appear to be that the latter text must provide an additional way of viewing the "possibility of experience" that allows us now to see a way in which there is included amongst "experience" the working of moral law (in the shape of the categorical imperative). Since the will belongs also to the world of understanding it would seem that Kant views it as the basis for providing laws that have to be added to those laid out in the Critique of Pure Reason.
The second sub-part of the fifth claim is where Allison finds the "crux" of the argument. Here Kant includes the claim that the being of sense is nonetheless subject to the laws of understanding which Allison argues are intended to be seen as holding simultaneously with the laws of the type described in the Critique of Pure Reason. This is followed by a paragraph that Allison sees as "explaining" the deduction apparently now given. It opens by declaring that categorical imperatives are "possible" but Allison views this claim as only being that we have had provided to us a reason for thinking that the categorical imperative is a necessary and not a sufficient condition of the argument that the idea of freedom makes me a member of an intelligible world. Further it is the "idea" of freedom only that makes one a member of such a world. Being a member of such a world is seen by Allison as having the mere intellectual capacity to think the idea of such freedom.
However Allison's conception of the "explanation" of the deduction runs into a problem with the appearance of a "sollen" (or "ought") that arises next when Kant states that I "ought" to conform my actions to the laws of the intelligible world. Allison thinks this move is here introduced illegitimately since all Kant has shown are epistemological grounds for being in two "worlds" and that a normative requirement does not arise from this alone. Part of the problem here is that Kant had earlier suggested that entirely belonging to the world of understanding would have produced the outcome that one simply "would" have followed the laws of understanding, a point that appears to render these laws, as Allison says, "vacuous". Allison argues that the problem with the text at this point arises from Kant not yet having explicitly to hand the distinction developed later in the Religion between two aspects of the "will".
Kant next provides an account that is meant to give a basis for "confirmation" of the deduction and involves the awakening of a "scoundrel" to consciousness of the moral law. The claim of how this takes place in Groundwork III is, however, contrasted negatively with the parallel discussion in the Critique of Practical Reason by Allison. In Groundwork III Kant appears to state that it is the idea of freedom alone which suffices to change the standpoint of the scoundrel. By contrast, in the Critique of Practical Reason the "fact of reason" is invoked to provide the basis for a view that one does stand independent of determination by sensuous needs. This passage, even so amended, is, however, less plausible as a piece of moral psychology than Kant provides later in the Religion.
When Allison turns to assessing the deduction as he has reconstructed it he begins with the question of how to understand the argument that the world of "understanding" provides a ground for that of sense that enables a normative claim to be made on behalf of the former. Key to this is the point that the difference between the two is centred in the argument of Groundwork III on the self alone. This enables Allison to make the point that we need not view the supremacy of the "world" of understanding as a claim about a specific "realm" separate from that of sense. Further at Ak. 4: 458 Kant speaks of the distinction between the two "worlds" as that of a change in "standpoint". This change requires us to view ourselves as capable of legislating laws and having the capacity to recognise our own legislation as binding upon us. When we look at the "two standpoints" practically Allison adds we can give the "law" of understanding a content by means of the moral law. Such "laws" can also then be understood as "reasons".
Allison also looks at the question of whether there is some specific non-empirical "interest" which supports the moral law and admits that it is impossible to account for the possibility of one. So Allison does not think Kant resolves or is capable of resolving this problem. However this leads to the question of the boundary of practical philosophy since this boundary is reached in terms of arguing that it is impossible to demonstrate the impossibility of freedom. Freedom is only an idea, not an "absolute reality" and it rests, for Allison, on viewing the relationship between a subject and an action in divergent ways. Part of Kant's defence of this claim was that it is essentially recognised within common human reason itself. The world of understanding, considered from a practical point of view, is the world of the will and this point requires us, given the general picture of the world, to see the will as something non-empirical. This supports the point that the "standpoint" of freedom is one that reason takes up in order to be able to think of itself practically.
However when Allison looks at the final passages of Groundwork III he argues that Kant here weakens some of the claims made earlier in the same section. Kant here states that the question of the possibility of the categorical imperative can be answered "to the extent that one can state the presupposition on which alone it is possible" (Ak. 4: 461). This "presupposition" is the idea of freedom. Allison's point is that this claim involves accepting that the deduction provides only a necessary condition of the possibility of the categorical imperative not conditions that are sufficient to account for its actual possibility. Thus it would appear, on this view, that the argument of Groundwork III is of less moment than it initially appears to be. This points, for Allison, to the superiority of the account of the "fact of reason" in the Critique of Practical Reason. Indeed the commentary Allison provides to the Groundwork ends with the claim that this work is "in many ways a transitional as well as a foundational work".
Allison's account of the third part of the Groundwork in his commentary on it is instructive as the deflationary view of transcendental idealism that is Allison's hallmark is used here primarily to buttress the sense that the transcendental distinction has a primarily practical significance. However the conception that arises of the argument of the third part of the Groundwork is certainly a disappointing one and worth contrast in subsequent postings with quite different readings to that of Allison.