Rawls on Autonomy

In 1980, almost a decade after publishing A Theory of Justice,  Rawls delivered three lectures on the topic of "Kantian Constructivism in Moral Theory" which were subsequently published in The Journal of Philosophy (from which page numbers are given). Over a series of postings I intend to present an exegesis of these lectures beginning here with the first which concerns the distinction between what Rawls terms "rational" and "full" autonomy. 

In opening the first lecture Rawls refers to the need to consider aspects of justice as fairness that he has not previously emphasised and to "set out more clearly the Kantian roots of that conception". This rationale for the lecture is set alongside a separate one of simply aiming to make the notion of Kantian constructivism more familiar. However the first rationale will be the more important given that what is being made familiar is integrally related to the further working out of the normative implications of A Theory of Justice.

The first part of the first lecture begins with a description of Kantian constructivism as specifying a conception of the person in a reasonable procedure of construction. Otherwise put the constructivist view being advanced sets up a procedure of construction that answers to certain requirements and the latter are understood to have their force in their reasonableness. Essentially the idea is to buttress the apparatus of Theory by showing how there is a connection between the person, as understood in a certain way, and the first principles of justice. The connection is by means of "a procedure of construction". 

Having begun by stating this point about the relation between the sense of the person and the first principles of justice Rawls goes on to argue that the conditions for justifying the latter require a public culture that is capable of sustaining them. In the absence of a sufficient basis for agreement amongst citizens the task of specifying a principle of justice becomes one of showing what is most reasonable to accept given a conception of the person. Within "democratic culture" as Rawls understands it here, which understanding is very broad, it is a sense that persons have moral capacities of a sort that is key. These moral capacities are what permit us to sustain the idea of treating persons as free and equal. If there is dispute at present over the way in which justice is centrally understood democratically then the Kantian procedure has to be one of demonstrating which principles of freedom and equality most reasonably meet the conditions of potential agreement.

Phrasing the beginning of his inquiry in terms of democratic culture ensures that Rawls conceives of his inquiry in a restricted way. It is not trans-historical but rather an inquiry concerned with modern conditions and it is not trans-cultural as it assumes the basic shape of democratic society. Further there is a form of hermeneutic at work in the inquiry being undertaken that expresses the view that there is at least a desire for agreement on the basic principles of justice. This desire for agreement is traced back here to the notion of "common sense" with Rawls assuming that it is either the case that we need to articulate the notions inherent within it or to propose to it instead conceptions and principles that are congenial to its "essential" form. As he further puts this:

The real task is to discover and formulate the deeper bases of agreement which one hopes are embedded in common sense, or even to originate and fashion starting points for common understanding by expressing in a new form the convictions found in the historical tradition by connecting them with a wide range of people's considered convictions: those which stand up to critical reflection. (518)

The way the process of disinterring from common sense the resolution of the problem modern democratic culture is apparently faced with is by means of articulation, in the first instance, of a conception of the person that is affirmed within this culture or that will be acceptable to those formed by it once critical reflection on it has taken place.

So Rawls is concerned with a "public conception" of justice that can be affirmed by all who regard their person in a certain way. This idea is part of the way that Rawls now understands the notion of "congruence" as what has to be provided is a conception of justice that fits our "deeper understanding of ourselves and our aspirations". This entails then that it fits the way that moral psychology resonates with us. This is reinforced by the view that the history and traditions embedded in this public life are resonant with the reasonable doctrine that is so uncovered. This does have a radical implication: "Apart from the procedure of constructing the principles of justice, there are no moral facts" (519).

After introducing the topic like this Rawls next specifies the way in which the conflict within democratic culture is understood by him although the ways he characterises this conflict is not singular. It is between two different traditions and one of them is associated with Locke, the other with Rousseau. However Rawls also refers to the difference Benjamin Constant spoke of between the liberties of the moderns and the liberties of the ancients with Locke associated with the moderns (and thus Rousseau with the ancients). The modern Lockean conception is concerned with civil liberties especially liberties of thought and conscience and also property rights and rights of association. The other ancient Rousseauist conception begins instead from equal political liberties and views civil liberties as subordinate. This appears to imply, in Kant's own terms, that Locke essentially provides the rudiments of a doctrine of private right whilst Rousseau provides one of public right and the task is bringing them together (as Kant sought to do in the Doctrine of Right).

Now in working out a way of specifying notions of freedom and equality that bring together these traditions Rawls introduces the idea of "model-conceptions" as embedded in his notion of justice as fairness. The two key notions in question are those of the "well-ordered society" and a "moral person". The "original position" is now presented as a third model-conception that mediates between the other two (and hence provides us with the means of understanding the process of construction). The means by which the "original position" works is by showing ways in which the citizens of a well-ordered society would have, using the capacities of moral persons, selected the principles that created their own society.

In the original position we begin, however, with a more restricted sense of the persons within it as being only rationally autonomous whereas the citizens of the well-ordered society would be fully autonomous. Rationally autonomous agents are those who possess the powers of formulation of hypothetical imperatives. Before proceeding further with a discussion of rational autonomy, however, Rawls first proceeds to describe the well-ordered society. Such a society would be regulated by a public conception of justice in the sense formulated by his account of institutions in the first section of Theory. This conception would be one understood to be accepted by all and known to be so shared. Further the basic structure of the society would be taken to actualize the conception of justice so understood and would be founded on reasonable beliefs. Within this arrangement the members of the society would treat each other as free and equal moral persons.

Moral persons would all possess the capacity to have a sense of justice and would have an equal right to determine the first principles of justice on which their society was governed. The freedom of these persons would consist in their ability to make claims on the design of institutions in the name of their fundamental aims and higher-order interests. The view of the well-ordered society is used to provide constraints on the way the original position should be set up as the idea that persons are free and equal has to be given place within the original position. In so beginning it operates fairly between persons and thus we have the idea that the procedure in question will settle the principles of justice fairly, an idea that is captured in the notion of justice as fairness.

Rawls' means of picturing the original condition is augmented, as in Theory, by the use of the device of the veil of ignorance. It is intended that the introduction of this device will ensure fairness between persons and thereby provide what Rawls terms "pure procedural justice". The principles of justice are to be constructed by a process of deliberation in which no antecedently given principles are given any particular weight. It is clear though that if a principle weighs heavily against rational assessment of interests it would be rejected. If the citizens of a well-ordered society are such as would regard themselves as moral persons then the parties to the original position must be such as to enable this self-conception to arise.

Having stated this much Rawls describes the capacities he takes to be essential to having moral powers. These are the capacity for an effective sense of justice on the one hand and the capacity to form and revise a rational conception of the good on the other. Persons have higher-order interests to realize and exercise these powers. These interests are then the ones that supremely have to be regulative of how the operation of the original condition is defined. We take the parties also to be "developed" morally in the sense that they aim at particular conceptions of the good. There is also a higher-order interest that each has of being able to prosecute this but this interest is not highest-order like that of realizing and exercising moral powers.

For the veil of ignorance not to prevent recognition of the powers so described we have to specify a means by which rational agreement can plausibly be reached. This is by means of an account of primary goods which will provide a yard-stick by means of which the parties to the original position can evaluate conceptions of justice. The primary goods are described here as including basic liberties (of Lockean sort), freedom of movement, powers and prerogatives of offices, incomes and wealth as all-purpose means and the social bases of self-respect. Primary goods are in general singled out as those generally necessary all-purpose means to enable us to realize and exercise our moral powers. The specified highest-order interests of persons thus select what is to count as a primary good.

So when we state that conceptions that would fail to recognise the interests of the parties to the agreement would fall, we are referring not to material interests, but rather to the interest they all have in enabling each other to develop and protect their moral powers. In the original position, therefore, the parties to the agreement are autonomous in not being bound beforehand by any given conception of justice but also in being moved solely by their highest order interests. This is what Rawls understands as the "rational autonomy" of the parties.

Full autonomy, by contrast, requires a further specification of social cooperation. Rawls argues that the notion of social cooperation contains two elements. The first is a sense of the fair terms of cooperation and these include conditions of reciprocity and mutuality. This element of cooperation is what Rawls now terms the Reasonable. The other element of cooperation, by contrast, is what he calls the Rational. The rational element of social cooperation concerns what each party to the cooperation in question is hoping to achieve by means of this cooperation. In the original position we have determined the rational by means of higher-order interests and acting rationally in relation to these interests includes adopting principles of instrumental reason. The Reasonable, by contrast, is incorporated into the background setup of the original position including in the sense that this position has to meet conditions of publicity.

The original position is set up to represent the minimum adequate notions of moral personality and when to this is added the sense equals in all relevant respects are to be treated equally this idea is meant to ensure that the original position is described in a way that is fair to all members of the agreement. The first subject of justice is also to be the basic structure of society or its main social institutions. 

The summation of the remarks Rawls presents here about the Reasonable and the Rational are to the effect that the Reasonable presupposes and subordinates the Rational. The Reasonable defines the fair terms of cooperation but it presupposes the Rational as without a sense that there are distinct conceptions of the good there is no real point to such cooperation. It restates essentially the argument made in Theory for the priority of the right over the good.

Full autonomy is a moral ideal that is part of the comprehensive ideal of the well-ordered society. Rational autonomy, by contrast, is the device by means of which the conception of the person is related to the procedure of arriving at definite principles of justice. 

Parfit and Kant's "Contractualist Formula"

The third of Parfit's 2002 lectures concerns contractualism and involves an extended discussion of what he calls Kant's "contractualist formula". Since the formula in question is not one that Kant himself specifically states and since Parfit's derivation of this formula occurs in the concluding part of the second of the 2002 lectures it is necessary to return to this derivation before the argument of the third lecture can be considered. This is what I will do in this posting, which is devoted only to uncovering the argument by which Parfit arrives at and justifies this formula.


In the final section of the second 2002 lecture Parfit turns to the question of whether, in the course of this lecture, he can be said to have misinterpreted Kant's understanding of universal law. Whilst considering this possible objection Parfit points to the way in which Kant discusses, in the case of beneficence, the application of the law of nature formula by means of appeal not to what "everyone" could will but rather to the nature of the specific person's willing. Parfit also rejects the view that Kant's universal law test should be seen in terms of Rawlsian notions of the veil of ignorance. Parfit presents both Scanlon and Rawls as offering not interpretations of Kantian notions of universal law but as instead providing ways of revising it. This leads to Parfit presenting a complicated diagram of possible ways the universal law formula could be understood. Prominent amongst the possible understandings for Parfit is the treatment given by Scanlon which leads Scanlon to state that an act is wrong "unless everyone could rationally will that everyone acts in this way". 


However whilst Parfit is favourably impressed by Scanlon's formula he does not leave the matter there. Scanlon's formula is one that Parfit thinks is impartial but that it goes further than we have to since some acts are right without being ones that everyone could perform. This leads Parfit to revising Scanlon's formula so that it becomes, as Parfit terms it, the "formula of universally willed acts" which states: "an act is wrong unless it could be rationally willed by everyone". This formula is now understood to be equivalent to the first half of the Formula of Humanity on Parfit's construal of the latter, which he has earlier termed "the Consent Principle". On the basis of this we treat people only in ways that they could rationally consent to. The wider formula of this that has become the formula of universally willed acts now adds that an act is wrong unless everyone, if they had the choice, could rationally choose the act in question as one that would be done. 


However despite the argument having thus given us a formula that is at least loosely related to something Kant said Parfit is not finished yet. Parfit had earlier invoked the quite different standard that he termed Kant's "moral belief" formula which involved stating that an act is wrong unless we could rationally will it to be true that everyone believed it was permissible. This formula introduces a lot of complicated epistemic issues that Parfit does not really discuss but at this point all he does is argue that the so-called "moral belief" formula is one that can be converted into Scanlonian terms so that it becomes a claim to the effect that an act is wrong unless everyone could rationally will it to be true that everyone believed such acts permissible. This version of the "moral belief" formula interprets the claim about belief in a better way in a sense since it is no longer merely a claim about a given person's belief states and this is surely preferable as otherwise we seem mired in subjectivism. This version of the moral belief claim is meant to be understood as a way of addressing the problem of what general principles of action we could all will. 


Not only is the Scanlonian notion of moral belief a reply to a genuine Kantian question it also has the advantage, according to Parfit, of being suggested by Kant's formulas of autonomy and the realm of ends. The point about the formula of autonomy is that it involves a claim about rational beings giving themselves universal law through the maxims of their will. The understanding of this that Parfit wishes to champion is in terms of the Scanlonian version of "moral belief". The Scanlonian version is also termed by Parfit a "formula of universally willed moral beliefs" but it is shortened and given the name "Kant's Contractualist Formula" so that it becomes: "we ought to act on the principles whose universal acceptance everyone could rationally will". Parfit concludes the second lecture by contrasting this formula with a Rawlsian formula that involves, instead, reference to choice under conditions where we know nothing about the circumstances of others. The contrast between these formulas will motivate the structure of the third 2002 lecture. I am making no comment here on the argument Parfit has given as I wanted merely to draw it out as the basis on which the formula that Parfit consistently in the third lecture refers to as "Kant's" formula, despite not being stated anywhere by Kant, is arrived at.

Rawls on Autonomy and Social Union

The final chapter of A Theory of Justice is concerned with the congruence of the right and the good in a well-ordered society. Another way of expressing the same point is to say that Rawls here attempts to show the relationship between the notion of goodness as rationality and justice as fairness. The overall point is thus to arrive at the view that an effective sense of justice as recognised within the governing of institutions is part of the good that would be rationally endorsed by each one of us.


The first two sections of Chapter IX are concerned with autonomy on the one hand and the idea of social union on the other. In starting with an account of autonomy Rawls wishes to begin by responding to suspicion of the psychological roots of the sense of justice. The basic suspicion he considers is motivated by a kind of psychological reductionism which views the origin of the sense of justice in the morality of authority by which young children are directed as casting suspicion upon the subsequent development of a morality of principles. In response Rawls points out that in the well-ordered society the basic rudiments of education have been governed by application of the principles of justice to the situation of the child. So none of the ideals upheld in the educational process are based on exploiting particular weaknesses or on devaluing the specific potentials of the child in question. 


The Kantian interpretation of justice as fairness requires that action on the principles of justice be understood as autonomous action which means that the principles in question express conditions that are defined by the nature of free and rational beings. So the basic guideline for educational practices has to be training for autonomy. Rawls wishes as well to show that autonomy is compatible with objectivity and does this by recourse to the device of the original position. Autonomous principles are chosen there as such principles involve recognition of the contracting parties as free and equal rational beings. The "objectivity" of the principles is stated in the generality of their application which include a sense of them that means they are stated without partiality. So the principles that are chosen in the original position are objective in not including deference to any given persons or treating any particular principle as based only on alleged special considerations.


Part of the point of such an objective mode of appraisal of principles is precisely to remove the barriers to agreement that are attached to specific positions in ordinary non-ideal intercourse. Acceptance of the principles of right and justice enables relations to others to be defined in terms that are civic rather than merely personal. So the idea of the original position is meant to give sense to both the ideas of autonomy and objectivity. This is a specific kind of view of these values and one that has to defended against other views. So autonomy is not to be understood, as it is often is, simply as respect for the particular conscience of any given individual. Rawls' earlier treatment of civil disobedience already demonstrated this feature of his view, a feature that prevents simple dependence on "subjective" views of principles. A conscience can be well or poorly guided and it is poorly guided if it fails to manifest respect for the principles of justice which is why justified civil disobedience was justified precisely by its reference to such principles and not simply by appeal to conscience alone. It is not conscience that is thus respected in the valuation of autonomy, it is rather the personality of the person, their attribute, that is, of being a person that is respected. 


So the earlier account of civil disobedience spelled out in Chapter VI of Theory is to be evaluated as specifying in terms of partial compliance theory the ideal conception of autonomy that Rawls is laying out at the beginning of Chapter IX. This is important as it indicates that autonomy is not to be viewed, in terms of justice as fairness, as a mere reference to either conscience or, with this, to reliance upon the virtue of integrity. Appeal to integrity has, as Rawls puts it here, great appeal in times of moral uncertainty where it appears to define a bed-rock value. Whilst Rawls indicates that such virtues are part of the excellence of free persons they are far from socially sufficient precisely due to the way they define virtues without reference to a contentful sense of life. It is only by reference to a kind of conception of what life one should live that the virtues of integrity define anything but it is then the case that this conception gives the value rather than the integrity of commitment taken alone.


Rawls arrives next at the point that the two principles of justice define what he terms "an Archimedean point" for appraising institutions as well as the desires and aspirations that institutions generate. Because we can refer to these principles we do not need to define an ideal of society by reference to principles of organic unity or by some sense of a pre-lapsarian past. However the objections to the general theory that Rawls presents have tended to insist on a sense of communitarian value that it is argued the allegedly "individualistic" bias of his theory does not allow recognition of. In response, in section 79, Rawls turns to elaborating a conception of social union that is meant to show that the congruence of the right and the good can define a well-ordered society in such a way that this society would recognise the achievement of the good of community.


In setting out a conception of social union that is meant to capture this sense of the good of community Rawls begins by reminding us of the conditions of the original position. A central aspect of the original position was that the parties know that they are subject to the circumstances of justice. Amongst such circumstances are the plurality of conceptions of the good amongst participants in the original position. This ensures that there is a potential conflict amongst participants as well as a unity of interest. Having reminded us of this plurality of conceptions as part of the circumstances of justice Rawls goes on to discuss how the relation between this fact and the conflicting one of the unity of interest of members of the agreement can be recognised in distinct ways by different positions. In making this point Rawls thus intends to draw an important contrast between the conception of justice as fairness and a key competing position.


The central idea that governs the way the original position is viewed by justice as fairness is that we begin from "the weakest possible" assumptions. The conditions are defined in ways that are "simple and reasonable" but the collision between interests is assumed to be great initially with the result that the way to resolve the conflict requires a comprehensive theory. Having specified the conditions of the original position in this way Rawls now contrasts the conception of it that is embodied in justice as fairness with a key rival view. The rival view is one that takes the basic structure of society as defined by the conditions that are fed only in to the original position by justice as fairness. On this conception we arrive at a notion that Rawls terms that of a "private society". Key to such a notion of society is that the entities that comprise it (whether individual persons or social groups) are assumed to have ends that are either competing or independent but not, in any event, complementary. A second element of the notion of "private society" is that institutions are taken to have no special value in themselves so that engaging with institutions is really a burden only undertaken for the realisation of private ends. The good of others is never taken as a prime datum for any actor.


On the conception of "private society" it follows that division of advantages is simply a function of power and strategic positioning. It is not impossible that a fair outcome might arise within such a set-up though it would be fortunate if it does so. Public goods will be valued largely instrumentally and relations judged in terms of the prices put upon them as in a market mechanism. The stability of such a society will rest primarily on effective use of sanctions. Once the idea of such a society is specified it becomes apparent that the alleged "individualism" of justice as fairness is not of the sort that endorses private social arrangements. And this is shown in the commitment of justice as fairness to the conception of goodness as rationality. 


On the view of goodness as rationality it follows that there are shared final ends that are possessed by persons and that common institutions express these ends. One of the reasons why this view flows from the idea of goodness as rationality is that individual potentialities are always greater than can be expressed so that each of us has, of necessity, to limit themselves in terms of what talents we choose to exercise. It is only through social union that we can participate in the total sum of the realised natural assets of others. This occurs both over time and within each generation. There are many forms of social union from families and friendships up to organisations that prosecute particular aims of scientific and artistic relations. However Rawls uses the example of a game to draw out the basic characteristics of social unions. With a game there are four sorts of ends: there is the end of the aim of the game as defined by its rules; the end of the motives of the particular participants including desire for exercise, etc. motives which can vary in weight and intensity for each participant; the social purposes of the game which may be little attended to by participants but evident to a third-party observer; and the shared end of all involved that this should be a "good" game. The last point is one that can only be reached by general compliance with the rules (the aim of the game) and by the players all playing their best. So the game is, in the broad sense, a collective achievement.


The notion of the shared end with which Rawls concludes his analysis of a game does not imply that in social unions all wish for the same thing (any more than they do in a game since most games are competitive!). It is rather meant to define an agreed scheme of conduct so that each can relate to the others  by means of it. Games are a simple illustration of a general point about social unions and a well-ordered society is not merely one social union amongst others but is rather the type of the social union of social unions. This is because it includes two characteristic features that are central to it. On the one hand it is a shared final end of all members of society and on the other its institutional forms are taken to have intrinsic value. So it defines the two central forms of goodness: it is both finally good and intrinsically good.


The well-ordered society is finally good in defining a shared end of mutual cooperation which arises from every member of the society having an effective sense of justice. This sense of justice defines a regulative condition for all and thus defines finality in a moral sense. The intrinsic good of the institutions involved is a more complicated matter. Firstly, the Kantian interpretation of justice as fairness requires us to say that upholding just institutions is a good for each member of the society. But, further, the Aristotelian Principle holds also for institutional forms and shows that a just constitutional order provides a framework for all smaller and more closely defined social unions. Each person understands the first principles that govern the whole scheme and the plan of each is given a wider vista than it would otherwise possess. The regulative public intention is that the constitutional order should realise the principles of justice. The collective activity thus engaged in defines a form of Aristotelian Principle.


The next point is that the moral virtues, excellences of persons, would be displayed in the public life of a well-ordered society. This shows, as Rawls puts it, that "the collective activity of justice is the preeminent form of human flourishing". So the public realisation of justice would define the value of community, the point that, at the beginning of section 79, Rawls wished to show. This does not mean, as Rawls concludes the section by demonstrating, that the ideal social union would abolish the division of labour. One of the reasons why it would not is that the ideal social order would express our dependence on others not try to overcome this. Similarly it is not going to be possible that any individual become an exemplar of all virtues or talents. It is rather the collective activity of society that is to become the overall good of each of its members. So the just social union of social unions would define not an abolition of separation of labour but rather a way in which the activities of each can come to manifest themselves in ways that enrich all.

Allison and Kant on Autonomy (IV)

The previous posting on Chapter 9 of Allison's commentary on the Groundwork looked at his treatment of the various formulas of the categorical imperative and the type of defence he offers for the view that they are all "equivalent", at least in terms of extensionality. In this posting, by contrast, I'm going to look at how Allison treats Kant's contrast between autonomous and heteronomous principles of morality.


Allison opens his treatment of the contrast between autonomous and heteronomous principles of morality by looking at the three different ways Kant views autonomy. On Allison's reading Kant relates to the principle of autonomy as the principle of morality, as a formula of the categorical imperative and and as a property of the will, taking these to be three distinct characterisations. However, the key point for Allison is that it is because autonomy is a property of the will that it is the principle of morality and not because it is presented as one of the formulas of the categorical imperative. The reason for making this claim is that the property of autonomy is what enables it to be said that the will can bind itself without referring thereby to anything "objectively" required beyond its own willing. It is because, however, that autonomy is the supreme principle of morality that this principle gives unconditioned authority to the categorical imperative. Once these two characterisations are brought together, however, it is less obvious than Allison suggests that we can claim that the reference to the categorical imperative is not part of what makes the principle of autonomy the supreme principle of morality. After all, in the very name of the "categorical imperative" we have reference to the notion that it is unconditionally binding, something that supports my contrary suggestion that there is an intimate connection between the notion of the categorical imperative and the idea of autonomy.


Now, in turning directly to the contrast with heteronomy, it follows easily from Allison's description that the notion of heteronomy can be captured as requiring that the will is not the source of the moral law but is rather dependent for its relation to the moral law on something beyond itself. However, many readings of Kant's description of heteronomy view this as requiring that heteronomous principles of morality essentially depend on a hedonistic kind of motivation (something that prevents, however, the recognition of intellectualist forms of principle being seen as heteronomous). There are quite a few problems with this, even when the question is restricted to considering what is going on for Kant when reference to sensations is paramount in the choice of moral principle. For example, as Allison points out, the benevolent philanthropist of Groundwork I feels pleasure in being benevolent but this does not necessarily mean that it is only for the sake of the pleasure that accompanies the benevolent acts that they are being benevolent. After all, if they did not think the acts they performed were genuinely addressing cases of need, they would not perform them.


The basic problem with heteronomous principles is not found in reference to a presumed hedonistic motivation but rather in a false conception of the relationship between the law and the agent. In adopting a heteronomous principle the assumption is made that the decision to will the law is insufficient for actions governed by it to have moral worth, perhaps on the grounds of an assumed "anarchic" conception of such autonomous willing. Whatever the rationale in any case, in adopting a heteronomous principle of morality, the agent assumes that worth is given to action by something intrinsically external to the act of willing itself and thus cannot find the property that is distinctive to morality within willing. There are, as Allison goes on to point out, two distinct types of ground for heteronomy: empirical and rational. Empirical theories are not all reducible to hedonism since sentimentalist accounts of morality can take there to be delight in the adoption of maxims that are virtuous. However they are classified by Kant in relation to hedonism in a different way, namely, in viewing the increase of "happiness" in a general sense (even separately from reference to the one willing) as the central matter. 


Allison's discussion of the contrast between autonomy and heteronomy is brief and concentrates primarily on making the view of heteronomy more complicated than many other interpreters have suggested. However the conclusion of the chapter on autonomy does suggest a basic problem with Allison's overall approach which rests on a failure to relate the unconditionally binding character of autonomy to the unconditional requirement that is stated in the categorical imperative clearly enough. This is not contingent to Allison's presentation, however, but essential to it given that his interpretation cannot recognise the claim that the principle of autonomy is bound up with the very notion of the categorical imperative.

Allison and Kant on Autonomy (III)

My last posting on Allison concentrated on how, in Chapter 9 of his commentary on the Groundwork, Allison discusses the question of the relationship between the formulas of the categorical imperative, and, particularly, some questions that arise from his view that there is a second formula of universal law that emerges subsequent to the formula of autonomy. In this posting I am going to continue to look at the discussion of formulas of the categorical imperative in Chapter 9, focusing now on how Allison treats the question of their alleged equivalence.


The question of their equivalence is one that is approached by Allison initially in terms of extensional equivalence, which would involve the claim that the distinct formulas would yield the same results for the same cases. This is distinct from an intensional equivalence claim which would, by contrast, entail that the reason for the sameness of results would be because each of the formulas has the same fundamental rationale. The question of extensional equivalence is complicated by the problems of alleged counter-examples to the formulas, which, in the case of the formula of the law of nature, are generally false positives which are, on Allison's reading, not applicable to the Formula of Humanity. Since Allison's solution was to treat these two formulas as yielding different kinds of universality it appears difficult to see how, on his interpretation, a thesis of extensional equivalence of the formulas can be sustained.


In response Allison ventures what he terms a "complete construction interpretation" of the categorical imperative which  suggests that the different formulas are parts of the construction involved in arriving at the full sense of the categorical imperative. So, for example, the formula of universal law is not treated on this view as settled by its initial statement since Kant's later treatment is taken to enrich it. Each of the formulas is taken by Allison as presenting a vantage upon the categorical imperative. The claim that the formula of universal law is "equivalent" to the formula of humanity is understood as a claim about the reconstructed formula of universality that Allison takes to have followed from the statement of the formula of autonomy. The second, enlarged, sense of universalisability is related to as one that requires endorsement by any rational being and thus to be equivalent to taking rational beings as ends-in-themselves. However, it is also the case that this interpretation takes the fuller sense of Kant's reference to a universal law of nature to be one that requires understanding these laws as teleological.


Allison contrasts his account with that of Onora O'Neill who, in contrast to Allison, has argued not merely for extensional equivalence of the formulas but also for their intensional equivalence. O'Neill's interpretation takes the force of the formula of universal law to concern the conditions of agency at work in willing and thus argues that maxims that are ruled out by it are maxims that undercut such conditions of agency. However, O'Neill does take the "perspective" of the formulas of universal law and of humanity to be distinct. On O'Neill's view the formula of universal law is a formula from the perspective of agents who acknowledge that others are also agents and requires maxims to be adopted that relate to this recognition. By contrast, on her view, the formula of humanity involves the recognition that actions affect others and thus requires us to adopt maxims that will regulate actions with this in mind. Allison points out that O'Neill's account fails to address the formula of the law of nature, which is peculiar given that this formula is the one applied to consideration of maxims.


Allison's account, based as it is on his "complete construction" hypothesis, views the earlier formulas as more "primitive" than the later and in taking this tack rules out the possibility of intensional equivalence that O'Neill's interpretation is aimed at salvaging. Whilst this attempt on O'Neill's part is certainly an interesting element of her interpretation, the view of different "perspectives" she builds in can only involve the claim that the "same" reasons are involved in a very thin sense of "same". Allison's rejection of the view that autonomy can be involved prior to being directly stated does not seem plausible to me, not least because of the appearance of claims about respect as early as the Groundwork I discussion of universal law. Whilst there is something to be said about formulas being either explicit or implicit in their appeal to autonomy I don't agree with the reading that takes autonomy not to be involved in the earlier formulas. 

Allison and Kant on Autonomy (II)

In my last posting on Allison I looked at the discussion he provides in his commentary on the Groundwork of the "derivation" of Kant's Formula of Autonomy. In this posting I am going to look at the second part of Allison's discussion which focuses on the relations between Kant's formulas of the categorical imperative as these follow the initial statement Kant makes of the Formula of Autonomy.


At Ak. 4: 436 Kant describes three ways of representing the principle of morality as being "fundamentally only so many formulas of the very same law" and also states that one of the formulas in question unites the other two in itself. The claim that the three formulas are all variants on the same law is termed by Allison the "singularity thesis". A claim to this effect goes way back in Kant's lectures since he criticised Baumgarten for proposing the view that there are "many" principles in ethics, indicating that on Kant's own view there is only one true principle. However, Allison is nonetheless attentive to a claim that is added by Kant to the singularity thesis to the effect that there is a difference all the same between the formulas. The difference is one that Kant terms "subjective" rather than "objective". Kant also relates this "subjective" difference to the means by which an idea of reason can, by means of analogy, be brought nearer to intuition.


Allison indicates that he finds the reference Kant makes here to analogy to be obscure even though he recognises that it is connected to Kant's general procedure of schematizing laws (as in the typic of the Critique of Practical Reason). In explicating the reference to analogy Allison looks to the Prolegomena where cognition according to analogy is described by Kant as based on "a perfect similarity between two relations in wholly dissimilar things"  (Ak. 4 357-8). Allison suggests that the different formulas of the categorical imperative represent it according to different analogies. But this is difficult for Allison who does not find the Formula of Humanity to involve either an analogy with the categorical imperative or a typic. In responding to this problem Allison suggests that if the first analogy (at work in the formula of the law of nature) represents a connection between the categorical imperative and causal laws, the second, by contrast, relates to nature as embodying a teleological order, something Allison rather loosely relates to the conception of the end-in-itself in the Formula of Humanity. Even if this were to be accepted it remains unclear how the relations between the three formulas of the categorical imperative are to be grasped in relation to the requirement of coming closer to intuition.


Allison's next move is to look at the general claim that one formula unites the other two in itself. When making this claim Kant refers to the view that every maxim contains a form, a matter and a complete determination. The form is also subsequently connected to the general claim that maxims express universality. By contrast, Allison takes the reference to complete determination to refer to the harmonization of maxims with each other through the realm of ends (which relates to his earlier claim that here nature is being treated in a teleological way). So the formula of the realm of ends is taken by Allison to be the formula that unites the other two into itself.


However the next point that arises in Allison's treatment is the claim that Kant now mentions a new formula. This is the formula, expressed at Ak. 4: 436 which is referred to by Kant as "universal" and expressed as: "act according to that maxim which can at the same time make itself into a universal law". The reason why this should be taken as a "new" formula rather than as a way of showing that the formula of universal law, correctly understood, presupposes autonomy, is not immediately obvious. This is not quite the same as the view that Allison terms the "consensus" one which takes this formula to be equivalent to the formula of universal law as it requires an understanding of the "new" formula as a product of reflection upon the formula of universal law. Allison mentions a different view of the "universal formula" which has been presented by Allen Wood who takes this formula to be equivalent to the Formula of Autonomy (which does not, of course, cut against the interpretation I have suggested). However, Wood's reason for adopting this view is an odd one since he locates the basis for the claim that the Formula of Autonomy is to be primary only in a view of "human psychology" alone.


In contrast to Wood, Allison views the formula of autonomy as requiring a second, stronger, form of universalisability (being inter- rather than intra- subjective). However, since Allison thinks the second form of universalisability is made explicit in the Formula of Humanity and yet that the Formula of Humanity does not make an explicit appeal to universalisability he provides a ground for taking the "new" universal formula to unite the formula of universal law with the formula of humanity (which, however, we saw that a moment ago Allison took the formula of the realm of ends to do). 


Allison takes there to be a significant difference between the original formula of universal law and the "new" universal formula on the grounds that the first refers to "willing" universal laws whilst the second, by contrast, refers to "making" them. In making this claim he agrees with Allen Wood but Allison and Wood have quite different reasons for making the claim. Wood's view was that being able to will a maxim as a universal law provides only a test for the permissibility of the maxim and that this test really only relates to maxims taken one at a time. By contrast, on Wood's view, making maxims into universal law is concerned with our maxims viewed collectively (not individually) and requires the consistency of maxims to be taken into account. Allison, by contrast to Wood, does not take the difference to be one where one of the formulas is better than the other at tracking coherence on the grounds that a coherent combination of maxims has no obvious moral merit. 


Allison distinguishes between a number of the formulas in a different way. This is in terms of viewing the formula of universal law as a "meta-formula" or second-order formula which underlies and is expressed by the distinct formulas of the law of nature, the formula of humanity and the formula of autonomy (which latter Allison now equates with that of the formula of the realm of ends). However, after making this claim Allison goes on to add that the formula of universal law is only a "skeletal" form of the concept of the categorical imperative. The fully developed form of it is rather given in the "new" universal formula and this claim leads Allison to the view that there are effectively two separate forms of the formula of universal law, the one given when the categorical imperative is first introduced and the alternative one that is arrived at now, in the aftermath of the Formula of Autonomy.


The discussion in this section is thus elaborate, as is often the case when the relationship between the formulas of the categorical imperative is in question. What has thus far been lacking is any serious inquiry into the alleged difference between the two supposedly different formulas of universal law. Assuming this difference does exist there are three questions that could be asked concerning it, which are, is the difference tantamount to ruling out their extensional equivalence, does it effect any claim there might be to intensional equivalence, and does it have any practical import in reference to the consideration of maxims? 

Allison and Kant on Autonomy (I)

In my last posting on Allison's commentary on the Groundwork I looked at Chapter 8 which discussed the Formula of Humanity. I'm going to move on now to Chapter 9 where Allison makes a series of inter-related but distinct claims about the uses Kant makes of the Formula of Autonomy but Chapter 9 is, at least thus far, the richest and most intricate chapter of the book so my discussion of this will be over a series of postings, not all in one bite.


Chapter 9 opens with a reiteration of Allison's general view of Groundwork II as providing a complete "construction" of the categorical imperative where the term "construction" is defined by Allison to mean "an account of the necessary and sufficient conditions of its possibility". There are 3 conditions that Allison views Kant as having set for the categorical imperative: firstly, it must have a universal form, secondly, it must presuppose something of absolute value and, thirdly, it must command unconditionally (or be binding). The last of these is related by Allison to the Formula of Autonomy whilst the second element is related by him to the Formula of Humanity. This distinction between the two formulas is one that I expressed some scepticism about last time but in this posting I want to look more narrowly at how Allison discusses the process by which Kant arrives at the statement of the Formula of Autonomy.


This is the subject of the first part of Chapter 9 where Allison broadly discusses what we could term the "derivation" of the Formula of Autonomy. Allison states that this derivation is not a "model of clarity". It is stated at Ak. 4: 431 where Kant derives it from restating the formulas of universal law and humanity and suggesting that autonomy arises from some kind of combination of the first two. Allison addresses here the question of how this derivation is supposed to work, concentrating initially on how Kant refers at Ak. 4: 431 to "humanity" which is in terms of being "the subject of all ends". Allison indicates that this expression is ambiguous as it could refer either to "those for whom ends exist", that is, to end-setters, or to "the objects of those ends", that is, the same end-setters considered as self-standing ends.  The latter seems to me more obvious and is aligned by Allison with the point that the rational agents so considered would set a limiting condition on both the ends set and the means by which I select them.


Turning over to the understanding of universal law, Allison likewise introduces a distinction between two ways of understanding it, namely either in terms of intra- or inter- subjective universalizability. Intra-universalizability is now understood by Allison as a first sense of universalizability with inter-universalizability a second sense. The formula of the law of nature is related by Allison to the first sense but the introduction of rational beings in the Formula of Humanity is viewed by him as leading us to the second sense. It is this second sense that is combined with the notion of humanity discussed in the previous paragraph and taken thereby to lead us to the notion of autonomy.


The principle of autonomy involves the idea that the law is something that one gives to oneself and since it is so viewed some critics of Kant assume that this gives a kind of green light to amoralism. On Allison's construal of the derivation of autonomy this is ruled out since autonomy requires universal legislation limited by reference to respect for ends in themselves. This sense of maxims being adopted out of respect for the law introduces a clear reference to a respect requirement into Allison's account.


Kant also provides a reason for thinking that the Formula of Autonomy rules out dependence of maxims referring to "interest" as, in being autonomous, the will that wills the law becomes a "supreme lawgiver" (Ak. 4: 432). This point is indicated by Allison to depend on the sense that the universality of the maxims in question is of the second type.


There are three distinct features introduced according to Allison when Kant arrives at the explicit statement of the Formula of Autonomy. Firstly, there is another formula of the categorical imperative, and this formula gives us the newly expanded notion of universality. Secondly, we thereby learn something new about willing morally. Finally, it leads us to the concluding stage of Kant's account of rational agency in Groundwork II. Included in the last point is the connection Kant appears to draw between autonomy and the realm of ends. It seems that to consider oneself autonomously is of a piece with viewing oneself as part of a realm of ends where this realm is understood by Allison to mean "a systematic unification in which the members are ends in themselves".


The "ends" in question are not "private" ends in their content or, as Allison puts it, "it does not assume that they have any such ends in particular". This way of understanding the realm of ends draws Kant very close to the Rawlsian understanding of the original position and Allison clearly acknowledges this without drawing any serious consequences for what this might say about the process of "constructing" the categorical imperative, something about which Rawls had quite a bit to say. Allison does, however, distinguish two ways of belonging to the realm of ends, as ordinary members of it on the one hand and the supreme head of it on the other. The difference between these is whilst both make the laws of the realm of ends in the sense of legislating them, only the ordinary members are also subject to them by which he means that the supreme head is not subject to any constraint as it is "a fully independent being".


The opening part of Chapter 9 thus includes a rather weak account of the derivation of the Formula of Autonomy patently largely meant to justify the general claim that it is a formula distinct from that of humanity without addressing the questions raised in my previous posting. After this point has been made Allison effectively draws on a Rawlsian notion of construction without clarifying the consequences of this or exploring it seriously. So the opening moves of the chapter have some philosophical problems but, in making the case that he does, Allison does draw out well that the binding character of the law is connected to autonomy and this is a key point.

Allison and Kant on Humanity

In my last posting on Allison I looked at Chapter 7 of his commentary on the Groundwork that addressed Kant's formulas of universal law. In this posting I'm going to look at Chapter 8 where Allison moves on to a discussion of Kant's formula of humanity.


Allison opens the chapter by citing Kant's claim that "the will is a capacity to determine itself to act in conformity with [gemass] the representation of certain laws" and adds to it that what Kant here brings out is that rational beings set their own ends for themselves, a point that refers us forward to Kant's notion of autonomy although Allison, somewhat oddly, fails to draw this consequence. That the concept of moral obligation presupposes an end that is necessary in itself, is, however, something that Allison suggests Kant had a long commitment to, pre-dating even the composition of the Groundwork. However, the ends that are presupposed by the categorical imperative are taken by Allison to be ends only in a negative sense. What this means is that they are ends that are sources of constraint on the acts one can permissibly perform. 


After opening with these claims about ends Allison moves on to more careful distinctions within the Kantian discussion of ends. Firstly, Allison looks at the difference between "objective" and "subjective" ends for Kant aligning objective ends with motives and subjective ends with inclinations. However Allison is also careful to point out that by "objective" ends Kant does not merely mean a source of reasons for action but also ends that are given by pure reason. Any end that is capable of grounding a categorical imperative must, on Allison's view, meet two conditions. It must, firstly, be objective in the sense of being given by pure reason, and, secondly, it must have a self-standing character.


Now the "end" that Kant takes to meet these conditions is that of "humanity" so the key question concerns what the notion of "humanity" consists in for Kant. This question is complicated by the fact that, in different texts, Kant contrasts "humanity" on some occasions with "animality" and on others with "personality". Christine Korsgaard appears to take the relationship between "humanity" and "animality" to be the central one but also to view the distinguishing feature of humanity to consist in a rather minimal capacity, namely, that of purely being able to set ends. By contrast, Allen Wood, who frames the distinction between "humanity" and "personality" as the central one, sees the capacity to set ends "through reason" as what holds together the capacities of humanity. 


However Wood seems to take the reference to "reason" here to mean that Kant frames "humanity" in the Formula of Humanity in a way that involves no specific reference to morality (as fits his view that humanity is primarily a contrastive term with "personality" as the latter clearly does incorporate moral notions in itself). Wood's basic reason appears to be that the categorical imperative requires preserving and respecting rational nature in general, not just its moral function. However, as Allison rightly points out, this is beside the point since the issue is not what the categorical imperative requires but rather in virtue of what capacity "humanity" has the property of being an end in itself. Allison thinks that ambiguity on Kant's point might well have led Wood to adopt this view but it is not one that Allison can accept and his rejection of this view is one that I certainly endorse.


Another view of the "capacity" of humanity that Allison critically discusses is that of Richard Dean who identifies "humanity" with the good will. Dean's point is that Kant's argument leads to the conclusion that something is good in all circumstances only if its value is independent of inclination and that this is how Kant identifies the good will so the good will must be equivalent to the capacity of humanity. Allison regards this claim as based on a conflation of two distinct senses of end, namely "end" as something to be effected and "end" as something self-standing. The good will is an end to be effected (perhaps the ultimate such one) but it is not a self-standing end. 


After canvassing these views, Allison takes instead the claim that Kant is making about humanity to consist in the view that it is the capacity to be moral that is identified by Kant with "humanity" and uses citations from the Critique of Practical Reason to help make this point. The mere capacity to set ends is described by Kant in the Metaphysics of Morals as having only "extrinsic value" (Ak. 6: 434) which appears to rule out Korsgaard's minimal view. However, what is problematic in Allison's identification of humanity with the capacity for morality is surely that it leaves Kant's view as a highly general one (and leaves out reference to what makes this capacity possible, namely, autonomy).


After going through the discussion concerning what the capacity for humanity consists in Allison moves on to looking at Kant's derivation of the formula of humanity which, he alleges, occupies only two paragraphs of the Groundwork. Allison views this derivation as having two steps. The first is that the possibility of a categorical imperative presupposes the existence of something as an end in itself and then gives an argument for elimination concerning what that something must be. The second step is to argue that we have good reasons to regard rational agents as ends in themselves.


The argument for elimination that accompanies the first step rules out objects of inclination, inclinations themselves and what Kant terms "things". Allison's account of the inclinations themselves indicates that we don't have to accept Kant's apparently more extreme views of inclinations to accept this claim. However whilst Allison appears to find Kant's argument by elimination convincing he does not uncover a specific argument for viewing rational beings as the appropriate "something" in the first stage of the argument which is why he turns to the second stage in order to uncover this.


There are three parts to the second stage of Kant's derivation according to Allison. The first part is the claim that human beings necessarily represent their own existence in this way so that it is a subjective principle of action (based on Ak. 4: 429). One of the points Allison stresses in presenting this claim is that when one sets an end, one makes it one's own end. Further, to take oneself as an end is to make the consequential claim that one cannot be regarded as nothing more than a means to some other end. So the unconditioned worth that Kant was after in the "something" lies in the end-setter rather than in any particular end that they set for themselves.


The second part of the argument for Allison moves beyond the "subjective" claim of the first part as it claims that every rational being necessarily represent their existence in this way so that it is an "objective" principle (Ak. 4: 429). Allison views this as making a claim for the objective validity of my conception of myself as an end in itself. Allison takes this to introduce a normative import to the discussion that he did not find present previously though it is unclear to me where he finds this import since he still does not uncover the source of the claim to reside in autonomy. The third part of the derivation for Allison is the translation into imperatival form of the principle that rational nature exists as an end in itself. 


The final part of Allison's chapter turns to looking at Kant's uses of examples of application of the Formula of Humanity in the second part of the Groundwork. Since Allen Wood, amongst others, have argued that the Formula of Humanity has a status that is practically superior to that of the formulas of universal law, this is particularly interesting to do with regard to the Formula of Humanity. The same examples are treated with regard to this formula as were related to the formula of the universal law of nature previously and they are given in the same order.


Kant begins with the example of suicide and asks whether the maxim related to it could subsist together with the idea of humanity as an end in itself (Ak. 4: 429). In denying that these are compatible Kant points out that since human beings are not merely things they cannot be used merely as means, and this mere means requirement of the Formula of Humanity is here mentioned for the first time by Allison. Allison finds the application of this requirement in this case to be unclear since he is not sure what end humanity is here being used for. Further, Allison seems to think it is a reply to this argument of Kant's to say that if the self is the source of all value that it is also the source of its own value, a view that cuts against any serious understanding of autonomy and brings out again his failure in this chapter to consider it.


Allison's positive view is that Kant rules out the maxim underlying the case of suicide considered on the grounds that it involves an authorisation to withdraw from obligation, something that is the source of the "contradiction" Kant finds here and enables Allison to prescind from Kant's own reference to the mere means requirement. However the argument is not filled out by Allison and in leaving out the mere means requirement he seems not to support the positive sense of humanity very clearly here.


The case of false promising, by contrast, is one in which Allison can pull out the sense of the mere means requirement but only by conflating it with what Parfit has termed the "consent" requirement, namely by referring to the ability of the other to endorse the end embodied in the action. The discussion of neglecting one's talents appears not to refer to the mere means requirement to Allison but, given that he has seen the formula of humanity as giving only the basis for saying that certain ends should not be acted against and not a positive ground for any specific ends to be accomplished, he cannot see any basis for Kant's claim that reference to humanity should prevent neglect of talents. At least with regard to non-beneficence Allison's reading does partially work since the reference to ends one should not act against is here sufficient to strike down active non-beneficence but it is insufficient to produce a positive claim for beneficence.


In summary, then, Allison's chapter strikes me as problematic for two different reasons. Firstly, the identification of the capacity for humanity purely with a capacity for morality without identifying what this latter capacity consists in or what grounds it is insufficient. It ensures that the relationship between humanity and autonomy does not become clear. Secondly, and as a consequence of the first problem, Allison can give no persuasive account of the discussion of examples Kant gives in the second part of the Groundwork of the application of the Formula of Humanity. So whilst the discussion of the derivation of the Formula of Humanity is an intriguing addition to the secondary literature there are as many problems with Allison's account of humanity as he finds beset his predecessors views.

Henry Allison and Kant's *Groundwork*

Henry E. Allison has recently added to his already extensive collection of works on Kant a new work, Kant's Groundwork for the Metaphysics of Morals: A Commentary (Oxford University Press, 2011). Given the importance of Allison generally as an interpreter of Kant and the significance of the Groundwork for the understanding of Kant's contribution to moral philosophy the publication of this work must count as a signal moment in Kantian studies. Since this work is so significant I intend to treat it as a matter of serious import for this blog and will be responding to it with the same kind of detailed reactions I have developed in previous postings in regard to both Derek Parfit and John Rawls.


In this posting I will confine my attention to the "Introduction" to Allison's commentary. Allison here instantly makes the claim that the "main reason" why the Groundwork is particularly crucial for both understanding Kant's contribution to moral philosophy and for a view of modern moral philosophy as a whole, consists in the articulation in this work of a distinction between ethics that is based on "autonomy" and that which, by contrast, has a grounding in something heteronomous. In making this claim Allison is self-consciously following in the wake of J.B. Schneewind's historical work on modern moral philosophy but complementing the latter by seeing the emphasis on autonomy as a practical complement to Kant's alleged "Copernican revolution" in philosophy. However, Allison does acknowledge the point that emphasis on "autonomy" in Kant is connected to Rousseau's political conception that freedom involves obedience to a law one prescribes to oneself. Despite mentioning this point, however, Allison does not connect this either to Kant's later characterisation of enlightened reason or to Kant's articulation of the supreme principle of right in the Doctrine of Right preferring only to see a shift from Rousseau's political vocabulary to Kant's moral one, something that is not comprehensively accurate however appropriate it might be to the context of the Groundwork alone.


Unlike Rawls, however, Allison does not think that the effect of the emphasis on autonomy in Kant's ethics can be well described by means of a view that the right has priority over the good. In eschewing this presentation Allison prefers instead the formulation that the Kantian view involves the priority of the moral law over the good. However, again, the main reason for this apparent change in formulation is due to a distrust of Rawls' own conception of justice as fairness whereas Allison wants to emphasise the Kantian contribution to moral philosophy, something that suggests there is little substantive of importance in the altered formulation. 


Allison admits to the fact that serious study of the Groundwork has led to an alteration of views on his own part though he says little initially to indicate in what ways this has taken place. Similarly, whilst conceding that the days are, in my view thankfully, long past, when it could be held that the Groundwork could be regarded as constituting the main fabric of Kant's contribution to moral theory, Allison only modifies this old view by addressing the Groundwork as the place where Kantian meta-ethics is formulated whereas even this claim is perhaps more sweeping than is warranted. However, Allison is surely on safer ground when he reports, as a generally accepted view, that the categorical imperative is no longer treated as a kind of algorithm for morality by means of which considerations of moral judgment can be swept aside.


If this general point about the categorical imperative is taken to be significant however, it is only negatively so and would still leave almost everything left unsaid concerning it. Amongst other questions, in interpreting the Groundwork it is clearly important to arrive at a view of how it presents, as Allison puts it, "the nature and function of the categorical imperative" as well as treatments of the various alleged "derivations" it is subjected to in the text of the work and the supposed "problems" connected to its "deduction" in the third part of the work and Allison promises that his book will offer his own views on these key topics.


In addition to these promises Allison also makes clear that, for him, it is important to come to a view on the development of Kant's views from the statements in the Critique of Pure Reason on ethical topics, statements often not sufficiently treated. The point in particular on which Allison lays emphasis is that the Critique of Pure Reason appears to suggest that it has provided all that is needed in order for the ground to be laid for a metaphysics of morals so that there must have been some kind of change of view on this topic between 1781 and 1784.


Similarly, Allison indicates that amongst other historical considerations that his commentary will treat are responses to some of the key external influences on Kant, notably, the Wolffian "practical philosophy" to which Kant refers at one point in the Groundwork and Christian Garve's notion of "popular moral philosophy". The second section of the Groundwork is one he will be particularly treating as a riposte to Garve's views.


Allison's general account of the structure of the Groundwork is in terms of the work's focus on two objectives, firstly, searching for and, secondly, establishing, the supreme principle of morality. Essentially Groundwork I and II are taken to be concerned with the first of these tasks (what are conventionally described as the "derivation" of the categorical imperative) whilst Groundwork III, by contrast, treats of the third (generally termed the "deduction" of the categorical imperative). Allison also anticipates his general argument by indicating that on the latter, his overall verdict has not changed since the publication of his seminal work Kant's Theory of Freedom since he continues, apparently, to adopt the view argued for there, to the effect that the argument of Groundwork III is a failure though it remains to be seen if the nature of his treatment of the reasons for this alleged "failure" have remained the same.


An aspect of the treatment offered that is certainly surprising is that Allison also aspires to take seriously an objection that could be made to the structure of the Groundwork but that has not often been pressed in recent years. This is to the effect that the argument of the second section of the work is "redundant" given that Kant has already arrived at a statement of the categorical imperative by the close of the first part of the Groundwork (hence completing the first task of the work) whilst the third part tackles the task of "establishing" the categorical imperative (thus dealing with the work's second objective). Once put like this the argument as to why Kant introduces what is perhaps the most complex and sustained argument, the argument of Groundwork II, can appear perplexing. Further, whilst, as I said, the argument to this effect has been little advanced in recent years, it is articulated in a work by A.R.C. Duncan (a work of broadly intuitionist persuasion from the late 1950's) and has not been provided, as Allison suggests, with any kind of comprehensive response. Providing such a response is part of the rationale for Allison's commentary which ensures that addressing how well it succeeds in responding to this "redundancy argument" is part of seeing whether the work generally succeeds in adding something substantive to the great volume of works that already exist on the Groundwork.


It is within Groundwork II that Kant presents a number of different formulas of the categorical imperative though it has been a subject of sustained dispute exactly how many are there given and what the relationship between them consists in. Allison indicates that it will be part of his argument that the formulas represent "successive stages in the complete construction of the concept of the categorical imperative" and that they correlate with a progressive analysis of the concept of rational agency. Since the conception of the relationship of the categorical imperative to a picture of rational agency surely is central to the viability of the offering of the categorical imperative as truly being the supreme principle of morality this point is certainly one that is centrally important.


Allison's conception of Groundwork III is to the effect that within it Kant aims to demonstrate that our wills possess a specific property, namely that of autonomy. Unlike the Groundwork, however, Allison's books is divided into four parts, not three. The first part concerns "preliminaries", namely an analysis of the "preface" to the Groundwork and a treatment of Kant's responses to the projects of Garve and Wolff. The second part is a reading of Groundwork I, carried out over three chapters, addressing the understanding of the good will, maxims and the three "propositions" of this first part. The third part treats Groundwork II, looking at rational agency, the universal law formulas, the formula of humanity and an account of autonomy. Finally, the fourth part is a reading of Groundwork III, looking at the moral law, the presupposition of freedom and the "deduction" of the categorical imperative. All told the commentary promises to be perhaps the most comprehensive provided in English to date and to set a standard that will require response from all others addressing Kant's moral philosophy.

Skorupski on Autonomy and Practical Reason

After outlining his general view of "Critical Philosophy" the next stage of the introduction to Skorupski's The Domain of Reasons concerns practical rather than theoretical claims. In making this practical turn, however, Skorupski remains in contact with Kantian themes, beginning with an account of the place of autonomy in his view of reason.


Skorupski's conception of practical agency requires a correlate to the epistemic notion of self-audit and is found in the view of an agent grasping, from a first-person insight, grounds for action. Skorupski also relates epistemic self-audit to recognition of what Parfit pictures as intentional objectivity regarding some form of such recognition as part of what provides us with a warranted reason for action. Skorupski puts this by saying that what one has reason to believe depends on the set of facts one apperceives.


If knowledge of such apperceiving of facts takes place at all then there must be a similar warranted apperception of norms. This latter gives Skorupski his picture of practical autonomy but what distinguishes it from the view of autonomy that is central to Kant is that it is not pictured primarily in relation to moral norms but practical norms of any kind (hence taking hypothetical imperatives to have the same relation to autonomy as categorical ones). 


Skorupski's general conception of reason allows for evaluative reasons to be distinct from epistemic ones but he does not join Kant in picturing reason as a general ability. One of the bases of Skorupski's departure from Kant here is that he is committed to a radical kind of pluralism about the different kinds of reason though he still thinks that reason relations have an epistemic and ontological unity which arises from our autonomous capacities. Given this latter point it is not that clear why Skorupski denies validity to a general claim about "reason".


When it comes to practical reason in general Skorupski builds out from the notion of "warrant" indicating that norms give warranted reasons for action only when combined with warranted beliefs about one's desires. In making this claim Skorupski builds in a kind of sentimentalism which is part of what he means by referring to "evaluative" reasons. On Skorupski's view what there is reason to do can depend on what there is reason to feel, something that invokes his idea of the "Bridge principle" which is stated by him as follows:

Bridge says that there is reason for x to do what the affective responses that there is reason for x to feel would characteristically dispose x to do (24).

Part of the point of this is apparently to reduce the "gap" between reason and sentiment that Kant introduces though a lot here depends on the nature of the feelings invoked. Kant is far from outlawing reference to feeling in practical evaluation despite the popular picture (one encouraged to a degree by the argument of Groundwork I). But feeling is nonetheless distinguished for Kant between practical and pathological, a distinction that can also be understood to refer to a difference between justified action by reference to feeling and unjustified action by reference to feeling (with the latter requiring the subordination of reason to "passion"). It remains to be seen whether anything like this distinction can be built into Skorupski's account of practical reason.

Skorupski distinguishes between two distinct kinds of sentimentalism making clear that he is not a sentimentalist with regard to practical reason as this would entail that the Bridge principle was the only principle of practical reason. A basic reason why Skorupski is not a sentimentalist with regard to practical reason is that he admits an important place for appeals to impartiality. Skorupski also understands the agent-neutrality that is involved in universalist forms of morality (including Kant's) as requiring a form of impartiality. However, whilst Skorupski thinks that the need for some kind of appeal to impartiality is a requirement for a full theory of practical reason he departs from Kant in terms of not taking the categorical imperative to be the foundation (or "supreme principle") of morality. In assuming that universalism is effectively the same as impartiality Skorupski also makes a large step towards equating the Kantian view of universality with a consequentalist one.

Skorupski's sentimentalism is partially justified by his account of blame where blame is understood as a sentiment that it is right to feel. So Skorupski, whilst denying sentimentalism concerning practical reason adopts it with regard to morality. Again this is partly due to his view that the type of impartial appraisal he wishes to allow for concerns a relation to the Good (rather than the Right). 


Skorupski's departures from Kant in the area of practical reason thus are partly due to a failure on Skorupski's part to see the place of feeling in Kant's view of practical reason and partly due to a simple conflation of universalisation with impartiality (and of the latter with an implicit consequentialism). But it is also true that Skorupski wishes to motivate a conception of feelings as having an immanent rationality that is specific to them, a rationality that is part of the sense of them as having an internal intentional objectivity. So Skorupski wishes to extend the area of spontaneity so that it encompasses feeling but doesn't see that so doing requires an internal relation of feeling to reason as he thinks of reason as always immanent to the area in question.