Christine Korsgaard analyses the argument of Groundwork I in an article she originally published in 1989 and subsequently re-published both in her own collection of essays, Creating the Kingdom of Ends and also in a collection of essays on the Groundwork edited by Paul Guyer. Unlike the essay by Nelson Potter also published in Guyer's collection and which I responded to in a couple of previous postings, Korsgaard's account is less a step-by-step analysis of Kant's argument and more a reconstruction of its overall point set against an historico-philosophical view of certain tensions in moral philosophy.
Korsgaard sets the response to the argument of Groundwork I against the background of the dispute between internalists and externalists in contemporary moral philosophy though she promises to illuminate this argument by connecting it to an historical pre-decessor. The historical counterpart of this contemporary argument is presented as the conflict between sentimentalists and rationalists. In both cases there is a conflict concerning how to view what it is to be morally obligated. On Korsgaard's account obligation includes 2 elements: both the sense that we are motivated to act morally and that this motivation is one that is binding upon us. Essentially the argument between the sentimentalists and the rationalists concerned, on her view, a stress on one of these to the exclusion of the other. So the sentimentalists give an account of motivation that renders it non-necessary and hence insufficiently binding whilst the rationalists show it to be clearly binding but provide no serious view of its motivating power.
The point of focusing on this debate about obligation is that Kant identified obligation, as early as the "Prize Essay" as the "primary concept" of ethics. Further, on Korsgaard's account, the argument of Groundwork I is best seen as a description of what the concept of obligation involves. It requires seeing that a morally good action is done from the motive of duty or because it is understood to be right. However, for this analysis to be informative philosophically it is necessary to see it as a response to the previous debate between rationalists and sentimentalists.
This previous debate is characterised by Korsgaard through a description that primarily rests on Hume. Korsgaard sees Hume as giving a basis for the view that there is a split between moral obligation and moral binding force. Hume's argument is taken here from the Treatise on Human Nature where he suggests that it would be, to use an anachronistic phrase, a form of "empty formalism" to identify right action with action that has the right motive. The basis for Hume's claim is that we need first some information about what kind of thing is virtuous before we have regard to the virtue of right acts. This requires that we appeal to a moral sense that has to approve something other than the motive of duty in order to reach moral content. This creates a dilemma that Korsgaard traces through the early modern debate and that she states as follows: "If we retain the thesis that it is motives that essentially make actions right, it apparently must be motives other than a regard for rightness itself. On the other hand, if we are to retain the thesis that the primary motive of virtuous action is the motive of duty, we must have some way of identifying or defining right actions which does not depend on their motives".
Essentially the first solution is the one adopted by sentimentalists and leads them to appeal to natural affections such as sympathy. This kind of account can also be seen to underpin utilitarian views of motivation and to have reappeared in another form in the work of Amartya Sen. However it leads to two problems since the motives it appeals to appear only contingently connected to right action and thus to lack a binding character. Further, since the natural affections are the prime source of moral obligation the motive of obligation itself appears to be something only secondary (perhaps a kind of "back-up"). This conception of moral motivation also appears to view the motivation in question not only as extrinsically related to the moral obligation generated but also in terms of a spectator's conception of how it operates (thus taking it to be essentially passive in form).
By contrast, the rationalist view emphasises a sense of the active force of obligation and in this way retains the view that the motive of duty really does move us. However the rationalist views the motive of duty as one we are led to adopt by the terms of how things themselves are and requires us to see moral significance as actually in the things themselves, something that leads naturally to moral realism and intuitionism. Further it does little to provide an adequate account of how motivation itself operates.
By contrast to these pictures Kant's view in Groundwork I starts with the idea of the good will and then proceeds to analyse the notion of duty by which such a good will would be moved. This analysis articulates a distinction between motivations that refer to two types of inclination (direct as when we are moved by enjoyment and indirect by the hypothetical relation to something beyond the inclination itself) and contrasts these with motivation that is based on duty. Having set these points out Korsgaard reveals what is innovative in her account of the motive of duty which is to take this motive as having its worth in terms of the grounds on which a purpose gets adopted. This leads to her effectively identifying Kant's notion of "maxim" as a process whereby something is taken to be a normative reason. There are then distinct reasons why a purpose is regarded as worth having and it is the difference between them that determines distinct types of worth. As Korsgaard puts it: "Duty is not a different purpose, but a different ground for the adoption of a purpose".
The ground for adoption of a purpose that is specific to duty is that in its case a maxim is adopted because it possesses an intrinsic and not an extrinsic legal character (hence the sentimentalist stress on motivation is linked to the rationalist argument for binding force). The binding force of the maxim in this case turns on the universality of the form of the maxim relating to the grounds on which one can will a purpose. Korsgaard, in articulating this point by means of an example, refers to the case of the lying promise (Ak. 4: 402-3) which is used already in Groundwork I (something often forgotten). The argument here is presented as turning on the impossibility of willing a universal law of lying as such a law would "destroy itself". The point being, as Korsgaard puts it, that "he cannot rationally will to act on this maxim at the same time as he wills it as a law".
This claim that it is the will of the person themselves as reflected in the maxim adopted that requires them to see the impossibility of the world that such a will apparently aims for demonstrates the nature of the law in the universal law. This points already for Korsgaard to the emphasis on autonomy that is arguably only made explicit in Groundwork II.
Interestingly, Korsgaard's reconstruction of the point of the argument of Groundwork I says nothing about Kant's introduction of the notion of respect in this part of the work. For some thoughts on that see my previous posting on Nelson Potter.