Rawls on Construction and Objectivity

Rawls' third lecture on Kantian constructivism, as printed in The Journal of Philosophy, culminates the series he gave on this topic and addresses methodological contrasts between constructivism and some other views of moral theory. The central conception of Kantian constructivism has turned out to be the establishment of a connection between the first principles of justice and the conception of moral persons as possessed of freedom and equality. The connection is made by means of the procedure of construction by which rationally autonomous agents "subject to reasonable conditions" arrive at public principles. In concluding with a discussion of objectivity Rawls intends to secure the philosophical status of the constructivist procedure. This includes the basis for the view that principles of justice are best thought of as reasonable rather than true.

The first part of the third lecture consists in Rawls' summary of the work of Henry Sidgwick The Methods of Ethics. This work is the one that Rawls thinks of as being an "outstanding achievement" in moral theory. In making this claim Rawls also indicates what he takes "moral theory" to consist in: "the systematic and comparative study of moral conceptions" (554). Moral philosophy, by contrast, is wider since it takes its main question to be one of justification whether this is seen epistemologically or practically. Sidgwick's work is one Rawls thinks of as the "first truly academic work in moral theory" (554) which defines some of the comprehensive comparisons important to it. Sidgwick saw that moral theory was significant for moral philosophy but his work is seen as limited in at least two crucial ways by Rawls. The first key limitation in Sidgwick's approach according to Rawls is that he gives little attention to the conception of the person and the social role of morality. The reason for this is that Sidgwick assumed that a "method" of ethics is specified by first principles that aim at reaching true judgments. So justification is seen as a primarily epistemological problem.

The focus on truth does not only lead Sidgwick in this epistemic direction but it also leads him to restrict attention to first principles rather than to articulating views of the person and the social role of morality. Due to these elements of his view of "methods" of ethics Sidgwick does not articulate constructivism as a method. Sidgwick also failed to see that Kant presents us with a distinctive method of ethics since Sidgwick understood the categorical imperative as equivalent to a principle of equity that requires treating persons similarly in similar situations. Because Kant's view is not regarded as providing us with a "method" of ethics Sidgwick is left counting only egoism, intuitionism and utilitarianism as real methods says Rawls. In making this statement Rawls is wrong since he leaves out Sidgwick's view of common sense morality, something of cardinal significance in The Methods of Ethics. In claiming also simply that Sidgwick "favours" the utilitarian method Rawls also neglects here the large scope Sidgwick subsequently gives to intuitionist elements in his treatment. But Rawls is certainly right to claim that Kant is given no sustained analysis by Sidgwick who spends more time on replying to Kant's view of freedom than he does to any clear account of the central elements of Kant's moral theory.

After opening with this discussion of Sidgwick Rawls moves on to a contrast between Kantian constructivism as conceived in these lectures with rational intuitionism. Rational intuitionism is presented here as grounded on two views. These are, firstly, that the central moral concepts are not analyzable in nonmoral terms and, secondly, that first principles of morals are self-evident. So it follows from these claims that agreement in judgment would be based upon recognition of self-evident truths about good reasons with these reasons fixed by a moral order that is independent of and prior to our conception of the person. In characterising intuitionism in this way Rawls sees it as favouring a priority of the good over the right and capturing this priority in terms of the moral order being prior to the person. The contents of the doctrine that ensue can be varied for intuitionists and now he outlines a sense in which utilitarian views could even be said to follow from them. 

The point of discussing intuitionism is to indicate that heteronomous conceptions of morality need not be naturalistic. And heteronomy is now presented by Rawls as the claim that the basic truths of morality are independent of (or justifiable without reference to) a conception of the person.  A Kantian claim, which includes the notion of synthetic a priori judgments, is, by contrast, one that requires practical reason to include central formulations of the person. So first principles are related to this sense of the person with the idea of it being embedded in the notion of practical reason in some way. By contrast the conception of the person for an intuitionist is much more sparse. 

If the view of the person is one of the means by which intuitionism can be contrasted with Kantian constructivism the other that Rawls considers concerns the limitations of moral deliberations. Constructivists, as viewed by Rawls, do not view the powers of reflection and judgment involved in moral deliberation as fixed in any absolute way. Rather they are reflections of the public culture to which they belong. This is why justice as fairness can be presented through the device of the original position as arriving at public agreements. The conception of justice involved here is one that relates to the framework of deliberation that produces it rather than pre-existing such deliberation. The limitations of such deliberation are reflected also in the narrow arena within which clear agreement is defined as possible, for example, within justice as fairness, the basic structure. Requirements that relate to the subject of justice are also built into the scheme of justice as fairness such as the ability of principles to meet criteria of publicity. In order to meet such requirements principles cannot be above certain levels of complexity. Certain facts are thus rendered irrelevant in consideration of the constraints of the procedure adopted.

In principle also the procedure of constructivism requires as well priority rules of certain sorts such as the priority of justice over efficiency or that of the first principle of justice over the second. The plausibility of these priority rules enables ways of dealing with the complexity of the many prima facie rules that are generally stated. The idea of the basic structure also plays the special role of defining what background justice is understood to be like. Comparisons of what outcomes are better or worse are also enabled by the reference to the notion of primary goods. This latter notion requires some sense of the agents who are to be party to the agreement that can emerge from the original position.

Alongside these constraints of a constructivist procedure is the abandonment of moral truth as the basic pursuit of moral doctrines. No specific principles are assumed to be correct prior to the initiation of the procedure of construction and no moral facts are taken to hold separately from this procedure. The first principles single out what facts would count as reasons of justice. So the procedure of construction of the original position determines what "facts" are to be counted. Constructivism also requires no specific doctrine of truth. 

The main ideals of justice as fairness are found in the "model-conceptions" of the well ordered society and the moral person. The original position enables representation of the Reasonable and the Rational and from its constraints the principles of justice emerge for the citizens of a well ordered society. So the procedure of construction has an intimate link to the way that principles of justice appear to be embedded in conditions of our life by means of reflection on that life. These conditions are what Rawls described as the circumstances of justice and consideration of these circumstances as constrained by the sense of what is Reasonable produces, in relation to our "model conceptions" the sense of the principles of justice.

The Kantian view, as seen by Rawls, addressees the public culture of a democratic society in order to bring to awareness a conception of the person and social cooperation that is implicit within this culture. This does not mean our society is well ordered since the public conception of justice is disputed within it. But the procedure of construction is one that has to be congruent with elements of our existing culture in such a way that it can meet criteria of public justification. First principles that are presented by constructivism are not "true" but reasonable and what is rather "true" are statements about how derivative principles follow from first principles. 

The notion of rational intuition is one that Rawls wishes to argue is an unnecessary one for the attainment of the goal of moral philosophy. Objectivity does not require the standpoint of the universe as Sidgwick suggests. It requires instead a social point of view that is publicly shared or would be in a well ordered society. Such a point of view is what is required within a basic structure and which models the conception of the person within such a structure. The person should be seen not in epistemic but in practical terms, that is, as involved in social cooperation. The means by which this determination of the person operates is through the avenue of potential agreement on public conceptions. Such agreement emerges from a view of the person as possessed of moral powers that articulate a notion of autonomous reflection. It is by these means Rawls hopes to articulate a way past the impasse in contemporary societies about the appropriate way of viewing justice.

Systematic Overview of Rawls' *Theory of Justice*

Over the last year I have undertaken the task of providing a commentary on the whole of John Rawls' A Theory of Justice. The commentary began on 11th April 2011 and concluded on 11th June 2012 and the postings are mainly inter-linked. I suspect, though am willing to stand corrected, that this is the most extensive commentary on Theory current in the blogosphere. In this posting I want to step back from the details that have been the subject of the postings that were done as the commentary was on-going to reflect on Theory as a whole, giving a general overview of the book and indicating the ways in which the position it articulates holds together. 


A Theory of Justice is a work of nine chapters that is divided evenly into three parts. The titles of the three parts are not, however, particularly helpful in describing the contents of the three parts in question and nor is it the case that Rawls identifies securely the bifurcation of arguments that the work presents. Starting with the division of arguments first, it is important to identify that there are two essential strategies adopted in the work and which are meant to work together. The first and philosophically the most significant is the construction of the original position and the justification of the principles of justice and associated principles for individuals by means of this device. Included in the construction of the original position are formal constraints on the concept of right, a description of what are termed "the circumstances of justice" and a description of the conception of rationality that will enable us to derive the principles of justice. The central philosophical arguments of the book, that belong to Rawls' advocacy of "constructivism", are intended to be justified by means of the construction of the original position. However, whilst this is so, there are included within Theory a number of considerations that are not built from the arguments concerning the original position and hence are not "constructed" philosophically. These considerations are appealed to by reference to "intuition" or "common sense" and Rawls frequently marks them by indicating that accounts using them are not, in his favoured sense of the term, "arguments" strictly speaking. This does not mean, however, that reference to these considerations is without point or relevance within the overall structure of the book. Rather the considerations that are advanced by these means belong to Rawls' general justificatory strategy of showing that there are conditions that can be generally seen to hold or be agreed to and that the original position, in various ways, secures but which are not definitively established by means of it.


The discussion of Rawls' strategies of argument and justification can be illustrated and the nature of the work's contentions be better viewed if we now turn to the way the work is structured. I mentioned above that the titles of the parts are not very helpful in describing the content of what they deal with. The first part is broadly termed "Theory" and in some respects this does describe what is discussed in the first part of the book though only in some respects. The three chapters of the first part discuss, respectively: a) a first general and largely intuitive presentation of the main ideas of the whole theory of justice combined with an initial contrast of it to the standpoint of classical utilitarianism; b) an initial formulation of the two principles of justice and associated principles for individuals; c) an interpretation of the initial situation as the original position and a construction of the original position that culminates in a construction of the two principles of justice and a more substantive contrast of the contract view with classical utilitarianism and also with "average" utility.


It is evident from viewing the structure of the first part of the book that Rawls does not begin with his philosophically favoured method of construction by means of the original position as it is only with the third chapter that he even states this method and uses it to derive both the original position itself and the principles of justice as established by means of it. The first two chapters of the book thus draw upon intuitive considerations of common sense and are meant, by this means, to lead us to accepting the view that the construction of the original position is something we have good reason to undertake and that it will provide us with a more secure basis for the two principles of justice that are initially introduced without its use. Given that the "theory" of justice essentially is a theory that argues for the two principles as the basis of the best considered view of justice and bases this on the construction of the original position it follows that it is only the third chapter of the first part that really states the "Theory" that the whole first part is named after. The first two chapters would be better viewed as being "on the way" to theory rather than statements of the theory itself.


The first chapter of the work is clearly introductory and here we find some key conceptions that are not justified as yet but which will be central to the whole theory subsequently. These include the notion of the "basic structure" as the subject of justice, the idea of the well-ordered society, the original position itself and the problem of how to resolve the difficulty of determining the priority of different principles. The basis of introduction of "intuitionism" as a position is really only in terms of it being a kind of indication of the view that this problem is insoluble and thus a proposal of ad hoc ways of dealing with it. The introduction of the "priority problem" and the other aspects of considerations that are taken to be important to the theory of justice, so important indeed, that section 3 of the first chapter identifies already "the main idea" of the theory are all presented in ways that do not derive them directly from the device of the original position itself. What this means is that the central ideas of Chapter One emerge as Rawls puts it in the concluding section of the whole book from "the tradition of moral philosophy which comprises the historical consensus" about what is central. Arguments which take issue with some of the central ideas introduced within the first chapter (such as those Gerry Cohen used against the idea of the "basic structure") are thus ones that aim not at the central constructive procedure of Theory but rather against its inheritance of concepts.


The second chapter of the work builds on the basic ideas introduced in the first chapter and states, albeit in a preliminary and intuitive way, the two principles of justice and attendant principles for individuals, introducing as well the notion of "primary social goods". Included in this chapter is an argument that concerns the second principle in particular and gives a basis for the second part of the second principle being termed the "difference principle", a point that leads away from certain views of equality towards a very specific way that egalitarianism is to be considered. Again these arguments are not derived from the original position and are stated prior to its construction. Similarly the principle of fairness and the natural duties individuals owe to each other are drawn from intuitive considerations and are not thus here really justified philosophically even though there are some reasons given for favouring them.


It is with the third chapter that Rawls really arrives for the first time at his philosophically favoured strategy of arguing for the principles of justice by means of the construction of the original position. The construction of it proceeds by four devices: a) an outline of the alternative views of justice that will be considered by means of it; b) articulation of the formal constraints of the concept of right; c) use of the "veil of ignorance"; d) an account of the procedure of rational choice within it. Of these four elements the presentation of the alternatives is least secure since it cannot be justified through the original position that Rawls is in the process of constructing. Centrally the traditions that formed the nexus of the considerations of the first two chapters are at work in identifying the views that will be tested by means of the use of the device of the original position and hence reflect an acceptance of the "consensus" concerning the views to be taken seriously. Interestingly, although classical liberalism was considered in the second chapter discussion of the second part of the second principle of justice it is not treated here as providing a conception of justice that will be viewed as an alternative to that provided by the two principles of justice. Nor are libertarian, socialist or communist views of justice, should there be any such, regarded here as providing alternative fundamental principles that need to be viewed as competitors to the two principles of justice. The contrast is instead primarily with variant forms of utilitarianism.


The account of the formal constraints of right determines the construction of the original position by describing conditions that the principles of justice will have to meet. The requirements in question are described by Rawls as "natural enough" and said to be "suitably weak" eliminating in principle only egoistic conceptions that are viewed by Rawls as occupying only the status of the state of nature that would ensue were no agreement on principles of justice to be reached. The five conditions are not themselves constructed but rather constructive of the original position's means of determining the principles of justice that are worthy of consideration. Here we have conditions on such principles and they are five-fold: i) the principles should be general, an idea itself understood in "an intuitive fashion"; ii) they are to be "universal in application" to all who are moral persons which ensures that they have to have a certain simplicity and consistency and this element is derived from a "common basis" with generality; iii) publicity, a condition that is part of the idea of a contractarian standpoint and is said to be implied in the categorical imperative and gives a way of evaluating principles that is meant to support the stability of them; iv) provide an ordering on conflicting claims or resolve the intuitive "priority problem" that was mentioned in Chapter One; v) be final principles that are appealed to in practical reasoning or be the highest standards of such argument which shows that these principles will be over-riding in importance.


Looking at the formal constraints of right we can see that generality and universality are presented here as part of the construction of the original position despite themselves being only introduced in an intuitive way. Publicity is an idea that belongs to the very sense that it is useful to appeal to such a device as the original position so in a sense in accepting it as a criteria we do no more than take the original position seriously as a device. Similarly we have already accepted the importance of the priority problem so a resolution of it is something we have already agreed to take seriously, albeit on the intuitive grounds of its appearance from the consensus of previous forms of normative philosophy. Finally, that the principles should be over-riding is as much as to say that the procedure of construction of the original position is one that will be determinative for the principles of justice that can be seen to meet conditions of general agreement so this condition, like that of publicity, is part of the sense of accepting the device of the original position as seriously worthy of consideration. Given this review of the formal constraints of right we can see that, strictly speaking, publicity and finality emerge as the strongest arguments within the construction of the original position as they belong to the basic sense of it.


The "veil of ignorance" is a means by which the original position effects its general purpose of achieving conditions in which general agreement on principles of justice can be made. The "veil" becomes thicker as the construction of the original position progresses however as a thicker form of it is required to tackle the idea of "average" utility than is needed to respond to classical utilitarianism. The "veil" is also not so thick as to rule out an account of the "circumstances of justice". This is required since an account of these "circumstances" turns out to be part of what is meant by rational choice within the original position. These circumstances are themselves, however, an intuitive description of what Rousseau referred to as an account of "mean as they are" and pose essentially as what we might view as "realist constraints" that are separate from the constraints on the concept of right. They are "background conditions" as Rawls terms them that define necessities of life that have to be considered when the principles of justice are arrived at. Consideration of these circumstances is part of what secures the stability of the conception of justice. Included here are moderate scarcity and the sense that people have distinct conceptions of the good that often lead them to conflict with each other.


The result of the construction of the original position, a construction that defines the conditions of rational choice at work within it, is the basic argument of Chapter 3 for Rawls' two principles of justice, an argument that leads to a modification of the principles by contrast to their first intuitive presentation in Chapter 2. Chapter 3 concludes with a more general description of the requirements of classical utilitarianism that gives an idea of how the notion of impartiality that actuates it becomes one of impersonality, a contrast that allows for a first sense of the kind of persons that are engaged in making the rational choice that leads to the principles of justice being favoured, persons who are not, on Rawls' account, the "bare persons" of the impersonal utilitarian calculation, but, rather, the "determinate" persons we actually are. Interestingly this conception of "determinate" personhood arises as required by reference to the circumstances of justice.


The second part of the book concerns, so the title of it informs us, "institutions" and, again, whilst this title is not entirely apt, it is better than the title for the first part was. The three chapters of the second part are concerned with: a) a four-stage sequence for principles of institutions that are said to articulate the basic structure of society and related to a "Kantian Interpretation" of the doctrine of justice as fairness; b) a more extensive consideration of the second principle of justice that is related to a description of the functions of government; c) a construction of the principles for individuals that are related to the constitution of the basic structure of society and which lead to consideration of the special problem of civil disobedience. 


All three of the chapters of the second part thus do consider "institutions" in a sense though the way they do so requires careful refinement of the principles already given and further determination of these principles. The first of these chapters is officially centred on the idea of equal liberty but the idea of the four-stage sequence that is considered at the beginning of the chapter is said, again, to be an account of "our considered judgments", an intuitive idea. The four-stages involve an account of the justice of legislation, the constitutional arrangements for resolving the priority principle in practice, the grounds and limits of political obligation and the basis of judicial rules and following of them. The argument for the four stages involves an elaboration of the original position that allows for an ideal notion of constitutional formation that is a device for applying the principles of justice. Rawls subsequently goes on to define the conception of liberty more carefully and to consider the point that the "worth" of liberty to persons has to be also considered. This notion of the "worth" of liberty is also later presented as an account of its "fair value" to different persons and the basic structure is now determined as something that has to be arranged to "maximise the worth" to the least well off of the complete scheme of equal liberty, a point presented as defining "the end of social justice" (section 32).


In Chapter IV the first principle is given clearer determination and the rule for its priority over the second principle described. After Rawls has constructed the first principle and its priority he states the "Kantian Interpretation" of the overall doctrine of justice as fairness and makes clear that the original position is a way of rendering Kant's idea of the kingdom of ends. The veil of ignorance is defended here as a way of preventing heteronomy and the motivational assumptions within the original position are related to Kant's notion of autonomy. The original position is here defended as a way of replying to Sidgwick's problem with Kantian autonomy to the effect that the ground for the choice of moral principles is allegedly opaque for Kant. Thus the original position is something like a parallel to the third part of Kant's Groundwork or to his appeal to the "fact of reason". This demonstrates the centrality of the role of the original position in Rawls' theory.


If the fourth chapter defined the first principle of justice more clearly and worked through the means of application of it to institutional formations the fifth chapter applies a similar approach to the second principle of justice. In the process Rawls considers economic systems, justice between generations and a view of the elements of government. The process of consideration of these points leads to a sharper view of both the difference principle and the principle of fair equality of opportunity. The questions considered arise intuitively though the resolution of them involves appeal to the construction of the original position.


The fifth chapter constructs the two forms of principle for individuals that Rawls considers, the principles of natural duty and fairness. Having done so the substantive argument of the chapter considers a special problem that arises from non-ideal theory, namely, the problem of how to deal with "unjust laws" and the possibility of majority rule being unjust. This allows for a statement of a basic theory of civil disobedience. Whilst this is the only form of non-ideal theory considered within Theory it is justified by means of appeal to a problem that is defined within terms of near-perfect compliance and hence approximates to a form of a well-ordered society. 


The final part of the work is termed "ends" which suggests that what will be considered here is a general account of purposes but what is in fact at work here is a discussion of the theory of the good where Rawls moves from the "thin theory" that will be at work within the original position itself to a view of the "full theory" that would be articulated within a "well-ordered society". The three chapters of this final part consider: a) the need for a theory of the good and a basic account of the good for persons; b) the first part of the problem of stability which includes a discussion of moral psychology, the sense of justice and its basis; c) the second part of the problem of stability which includes a final argument for the priority of liberty and an account of how the good of persons is congruent with the social good of justice including a discussion of the unity of the self.


The third part of the work is particularly intricate and consideration of its role in Rawls' theory should include a sense of what kind of moral psychology is here being offered. It is a kind of ideal type of psychology that is normatively rather than "empirically" grounded though it is not intended that it should contravene the requirements for any empirical theory. The basic problem of the third part is to show that the theory of justice presents a stable conception in the sense that it would generate incentives within the members of the society it formed to maintain itself. 


The first chapter of this final part includes a basic theory of deliberative rationality and the introduction of what Rawls terms the "Aristotelian Principle" which latter is presented as meeting what a perfectionist should really want. This principle states a generic form of good for persons in terms of recognition of the complexity that is part of the general acceptance of what we all tend to take to be good. Interestingly this principle is not itself constructed but is meant to echo the consideration of circumstances of justice just as the requirements of deliberative rationality define a sense of self-regard that is meant to echo the requirements of right. In this chapter Rawls also describes the virtues as forms of excellence that relate to the rational notion of self-regard that would arise from seeing it in terms that echo the requirements of right. In a sense the arguments of this chapter belong to a fuller construction of the principles for individuals and fill them out in terms of a rudimentary theory of the virtues.


The second chapter of the third part presents, after giving a general idea of the well-ordered society (which is constructed) an ideal type of moral psychology in terms of how we would develop to the stage of accepting a principled relation to morality under ideal conditions. The point of this is to make psychologically realistic the view that there would be a sense of justice that had relative stability within a well-ordered society.


The third and final chapter of this part and of the book as a whole presents some reasons for thinking that destructive psychological propensities would have little hold within a well-ordered society but this argument is largely intuitive in form. By contrast the congruence argument (in relation to the compatibility of a person's good with the social sense of justice) is one that arises after a final account of the priority of liberty and an argument has been given against "dominant ends" conceptions of the good. The final congruence argument draws upon the argument against "dominant end" conceptions of the good as it shows that the view that the good of persons is congruent with the social good of justice depends upon acceptance of the social good as a defining constraint upon one's good, something reasonable given that there could be no "dominant end" for us as individuals. Given that the congruence argument works like this it follows that its account of the relation between the right and the good depends upon a sense that the structure of the good is one that, even in a thin sense, would not be such that we could find enough determinacy in it to resist the constraints of the right. So it is a form of constructive argument but one that is also related to the ideal type of moral psychology defended in Chapter 8.


Taking the book as a whole then the argument for the construction of the principles of justice in the first part is complemented by a description of the kinds of institutions that would have just form accompanied by a view of the kinds of persons that could sustain them. As a whole it requires the sense that the person in question could themselves be constructed to have the motivations that would sustain the basic structure given that this structure would itself provide them with enough of value and worth to make this plausible. Therefore the theory is as much a theory of the good as it is a theory of the right albeit a theory of the right that constrains the theory of the good.


In future postings I will review the way Rawls' work developed after Theory and also look at the kinds of criticism that Theory has received from others in order to combine these perspectives to understand the subsequent turns his work took. The point of this will be to enable an assessment of Theory that goes beyond viewing it in its own terms in order to see the way it fits into Rawls' overall work and how its central contentions and devices have fared within political philosophy generally. Most important for consideration here, however, will be the fate of the specifically Kantian elements of Rawls' account and the way they are bolstered and weakened at different stages of it.

Rawls and Common Sense Precepts of Justice

In my last posting on Rawls I discussed the form the two principles of justice reach in Chapter V of A Theory of Justice. Having stated this form I then wondered what justified the rest of the work, or at least, the rest of Chapter V. Chapter V is, in fact, subsequently, an argument concerning different views of justice to that articulated in the notion of justice as fairness. Two sections concern "common sense" precepts, one the notion of "mixed conceptions" of justice and a final one, the principle of perfection. In this posting I'm simply going to fix on the discussion of "common sense" precepts which is the main subject of sections 47 and 48 of Chapter V.


Section 47 opens with a reiteration of the point that the sketch of the system of institutions that satisfies the two principles of justice is now complete. Assuming the basic structure has this form the distribution that results will be just (or, minimally, not unjust) and the analogy here, as is typical of much of the argument of Theory, is with the outcome of a fair game. However, at this point, Rawls turns to consider the question of how the conception of justice as fairness relates to our common sense precepts of justice. 


Common sense precepts have within them the problem of priority that Rawls articulated back in Chapter I as arising for intuitionist conceptions and which the various arguments for priority of principles that he gave in the previous sections of Chapter V show a basis for having overcome on the conception of justice as fairness. The two principles of justice are taken to define the basic criterion of how we will understand justice but are themselves arranged in an hierarchical order and contain (in the case of the second principle) sub-parts that themselves need articulating in order.


Rawls now considers various types of "common sense" precept and how they arise. However, he is less interested in these precepts themselves than with how different conceptions of justice lead to different ways of weighing them. The difference between different conceptions of justice is not understood by Rawls to reside in different types of common sense precepts but, instead, in different types of attention to them. A society that provides for fair equality of opportunity (the first part of the second principle of justice) responds differently to common sense precepts once this principle is part of its basic structure than does another society that does not provide for fair equality of opportunity in this way.


For one thing the provision of opportunities according to this principle tends to produce greater levelling of incomes with the result that the precept of rewarding each according to training is given less weight in this society than in alternative ones that have not built the principle of fair equality of opportunity into their basic structure. What this example is meant to show is that the type of generality that common sense precepts possess is the wrong sort for the articulation of principles of justice. This is despite the fact that some common sense precepts do initially appear general enough, such as ones that stress the natural right of property in the fruits of our labour (a kind of Lockean view). However, for this view to really be generally applicable it must be the case that the distribution it refers to is part of a generally just order, it cannot simply define one. It is, after all, just one of many precepts with another, also often appealed to and clearly off-setting this, referring to distribution according to need.


In market situations the general problem, at work in the kind of Lockean conception, is one in which contributions are expected to be rewarded. But what matters for any given kind of contribution is the relation of it to the whole system of norms in question. In this respect the questions of choice of occupation and free association are parts of a system of justice but the overall principle of fair equality of opportunity is one that determines the rewards that would arise in relation to them.


If section 47 articulates the problem that common sense precepts lack sufficient generality to stand in for principles of justice at the level of the basic structure, section 48, by contrast, tackles one specific type of principle that would threaten to reduce questions of justice to questions of virtue. This would be the adoption of a general principle that rewards should be adjusted by connection to virtue or moral worth shown (a kind of "republic of virtue"). Justice as fairness is in opposition to such a notion despite recognising the place of legitimate expectations within systems of justice.


The extent of a contribution someone has made at a given time is often determined impersonally by mechanisms such as markets that price the reward of a given effort in ways that may have little connection to the individual effort involved. This is a problem with viewing the conception of just distributive shares as a process of maximising returns by reference to conscientious effort. It also shows the difficulty of adopting such a principle as a public one. Moral worth, whether defined through conscientious effort or in some other way, is not a principle of distributive justice. Indeed, so little is this the case according to Rawls that there would be something grossly offensive about thinking of justice in this way as he suggests in the following striking comparison: "For a society to organize itself with the aim of rewarding moral desert as a first principle would be like having the institution of property in order to punish thieves".


The reason for this striking analogy concerns the distinction Rawls draws between a conception of reward determined by virtue and one that draws on the correct notion of legitimate expectations. In the latter case the expectations are what arise from doing things encouraged by existing arrangements and these would be best defined, on Rawls' view, by reference to the principle of fairness and the natural duty of justice. Institutions are bound to realise legitimate expectations they have encouraged in relation to these superordinate principles of justice. However, even when we have a system of justice that is governed by such principles there is no way of ensuring that conscientious effort or any other such type of moral worth would lead to higher rewards.


In this respect distributive justice is quite unlike retributive justice. If the purpose of the latter is to uphold basic natural duties through the provision of penalties attached to their violation, the purpose of distributive justice is only to ensure a generally just order and to reward individuals for carrying out generally valued social functions, so the variations in rewards should only be concerned with promoting the general ends of the basic structure, not the elevation of particular individuals in a parallel to the degradation suffered by those who have violated natural duties.


In sum then, Rawls concludes that most common sense precepts of justice are insufficiently general to be possible candidates in the original position for the status of general principles of justice and he specifically rules out views that would conceive of social justice as a wide form of reward for virtue since the latter view conflates the purposes of distributive justice with those of retributive justice.

Related articles

• Rawls and Distributive Justice (III) (kantinternational.blogspot.com)
• Joshua Cohen, Rawls and Occupying Philosophy (kantinternational.blogspot.com)
• Rawls, Just Savings and Priority Principles (kantinternational.blogspot.com)
• The Kantian Interpretation of 'Justice As Fairness' (kantinternational.blogspot.com)
• Rawls on the Rule of Law and the Priority of Liberty (kantinternational.blogspot.com)
• Rawls on Liberty and Constitutions (kantinternational.blogspot.com)
• Rawls, Classical Utilitarianism and Benevolence (kantinternational.blogspot.com)
• The Two Principles of Justice and the Original Position (kantinternational.blogspot.com)
• Rawls and "average utility" (kantinternational.blogspot.com)
• Rawls and Political Economy (kantinternational.blogspot.com)
• Rawls on Toleration and Equal Liberty (kantinternational.blogspot.com)
• Rawls on the Principle of Participation (kantinternational.blogspot.com)

Joshua Cohen, Rawls and Occupying Philosophy

I listened to this podcast the other day. In it Professor Joshua Cohen a leading Rawlsian philosopher aligns the general outlook of A Theory of Justice to the "occupy" movement that has been significant recently both in the US and the UK. Whilst Cohen's remarks are rather slanted towards the US situation and also signally fail to address the question of whether the later Rawlsian conception of "political liberalism" fits as comfortably with the "occupy" movement as he suggests, it is, nonetheless, a very good, clear interview that outlines the general conception of "justice as fairness" very well. 

The Kantian Interpretation of 'Justice As Fairness'

Chapter IV of A Theory of Justice concludes with the central section 40 that discusses the sense in which Rawls takes it to be the case that 'justice as fairness' can be given a 'Kantian interpretation'. There is a great deal worthy of comment in this very rich section, not least with regard to how it both explicates a lot of Rawls' view and raises some important points about the Kantian implications of it.


The section opens by relating the general view of justice Rawls has set out to the Kantian conception of autonomy and emphasizes this conception over the place of universality and generality in Kant's ethics. The downgrading of the importance of universal law follows from an acceptance on Rawls' part of the force of the Hegelian objection of the "triviality" of the focus on universal law (although Hegel is not explicitly here referred to). In some respects this is surprising given the account Rawls gives subsequently of the "categorical imperative procedure" and is certainly not in accord with the work later done by constructivist interpreters of Kant.


However, whilst the downgrading of emphasis on universal law has problematic elements in it, there is also one basis for it that is creditable to Rawls and this is his intention to pay attention to the full scope of Kant's ethical view, a point made in an important footnote where Rawls wishes to distinguish Kant's view from the utilitarian conception of universality and attempts to easily reconcile Kant with utilitarian considerations. In making this point Rawls points, in a way that until extremely recently was unusual, at Kant's later ethical writings, particularly the Metaphysics of Morals although he also mentions the moral significance of Kant's works on religion and even refers to the Critique of Judgment. In the course of making this comment Rawls states that there was, at the time of writing Theory, 'no commentary on Kant's theory as a whole", adding, 'perhaps it would prove impossible to write'. Without saying that the work has addressed all elements of Kant's theory, since I am now certainly aware of a number that it does not, I think I can, without undue immodesty, point now to my own book, Kant's Practical Philosophy as covering rather more ground than any of the works Rawls here cites and going at least someway to disproving his pessimism concerning the possibility of addressing the ramifications of the whole theory.


Returning to the main text of section 40, Rawls' substantive philosophical point is to articulate the sense in which Kant makes moral principles objects of rational choice so that moral philosophy becomes the study of "the conception and outcome of a suitably defined rational decision". One of the points that Rawls derives from this is that understanding the point of legislation in relation to the kingdom of ends leads as a matter of course to the centrality of public principles although he neglects here to discuss the differing formulas of publicity Kant offers in Perpetual Peace. Free and equal rational beings are also presupposed in Kant's moral legislation.


Rawls articulates the original position as a way of realising the legislation of the kingdom of ends. One of the supporting considerations for this philosophical interpretation of the reference to the kingdom of ends is that Kant wishes to exclude heteronomous motivation and the adoption of the "veil of ignorance" ensures that the methodology of the original position maps this constraint. Effectively the ruling out of heteronomy is understood to require what Rawls earlier referred to as the formal constraints of right and the "veil of ignorance" is a means to ensure these formal constraints are met.


However, as Rawls goes on to add, there are additional elements in his construction that Kant does not refer to. One of the central ones is that of the basic structure although in referring to this Rawls neglects to follow the admonition of his own footnote since he doesn't mention here the formation of the basis of a state of right in Kant's Doctrine of Right but he does indicate generally that when all Kant's ethical writings are taken together something like the basic structure can be seen to emerge.


The principles of justice are also suggested to be analogous to the categorical imperative again when the latter is viewed primarily as a constraint that applies to persons understood as purely free, equal and rational. Just as the exclusion of heteronomy is mirrored in the "veil of ignorance" so also is the adoption of specific ends as required in hypothetical imperatives. However, Rawls does retain the notion of primary goods although he does assume these arise only from the most general assumptions about rationality and human conditions. The assumption of motivations in the original condition was also one of general mutual disinterest and this is again mapped in relation to the Kantian conception of autonomy. 


Having made these points Rawls follows Sidgwick in identifying a problem with Kantian ethics, a 'problem' first articulated in the 'appendix' to the latter's Methods of Ethics. The 'problem' concerns an alleged ambiguity in Kant's view of freedom as on the one hand it is understood as a basis of choice whilst on the other it is also argued that choice of the law indicates a manifestation of freedom itself. This 'problem' has been examined at length since Rawls wrote this section of Theory with extensive accounts of the different types of freedom in Kant being undertaken by, for example, Henry Allison in his important book Kant's Theory of Freedom (amongst many other key treatments). It is a matter of distinguishing different ways in which Kant understands freedom that allows for the argument that "the scoundrel" as Rawls puts it does not realise freedom in the same sense as the one following the law and is far from being, as Rawls, following Sidgwick, thinks, a "decisive objection" to the Kantian view of freedom. Rawls says that Kant does not show "that acting from the moral law expresses our nature in identifiable ways that acting from contrary principles does not" but this is a mere assertion on Rawls' part since he does nothing to examine the evidence here.


It is likely far from being Rawls' case to engage Kant on this point since he wishes instead to articulate the original position as making good the alleged lack in Kant's view. Saying this does mean that the original position is taken to be an analogue to the noumenal view of choice with built-in constraints showing the ground for rational decision. In saying this however Rawls effectively reprises again the argument Kant gives not of moral choice in relation to autonomy but instead free choice in relation to right showing a surprising confusion of levels in Kantian theory afflicts Rawls' account. The description of the original position does not, as he claims, replicate that of noumenal freedom. Instead it replicates the contract position of right showing the ground for the restriction of external freedom being what can also realise it (the argument of his own previous two sections).


Rawls however is right to present Kant's view as one in which liberty is acting in accord with a law we give ourselves and that the understanding of shame, in Kant, is precisely one in which we fail to live up to such an ideal. The comprehension of Kant's view as an ethic of mutual respect is correct. The original position is presented in general by Rawls as a procedural interpretation of autonomy and the categorical imperative "within the framework of an empirical theory". One of the consequences of this is that Rawls does not investigate the notion that there is such a thing as pure practical reason. Another is that he conflates conditions of right with conditions of noumenal freedom. A third is that Kant's transcendental view is presented by Rawls as "transcendent", a mistake that is particularly striking. Finally, Rawls' view is explicitly distinguished from Kant's in the sense that whilst Kant's view applies to rational beings as such, Rawls' view is only meant to apply to human beings. The basic result of this is that Rawls gives up on providing a general theory of normativity or practical reason. One of the reasons for this is that Rawls wishes to give up on Kant's "dualisms", something that shows the philosophical modesty of Rawls' view. This modesty involves abstraction from central philosophical problems, an abstraction that much recent moral philosophy has, in my view rightly, given up on. The revival of more ambitious and comprehensive theories of practical reason is a return to the view of it that contemporary philosophers share with Kant and shows that the influence of Rawls on a central element of philosophy has waned.

Related articles

• Rawls on the Rule of Law and the Priority of Liberty (kantinternational.blogspot.com)
• Constructing the Original Position (kantinternational.blogspot.com)
• The Two Principles of Justice and the Original Position (kantinternational.blogspot.com)
• Rawls' Principles for Individuals (kantinternational.blogspot.com)
• From the Initial Situation to the Original Position (kantinternational.blogspot.com)
• Rawls, Classical Utilitarianism and Benevolence (kantinternational.blogspot.com)
• Nelson Potter and Groundwork I (kantinternational.blogspot.com)
• Rawls on Toleration and Equal Liberty (kantinternational.blogspot.com)
• Rawls and "average utility" (kantinternational.blogspot.com)
• Parfit, Sidgwick and Kant (kantinternational.blogspot.com)
• Choosing Principles in the Original Position (kantinternational.blogspot.com)
• The Difference Principle and Fair Equality of Opportunity (kantinternational.blogspot.com)

Rawls and "Classical Utilitarianism" (II)

In a recent posting I looked at Rawls' critical response to utilitarianism in the fifth section of the first chapter of A Theory of Justice. However, the reply to "classical utilitarianism" carries on after this fifth section into the sixth section. Whilst the discussion in the sixth section does repeat some of the motifs of the fifth section, particularly the claim concerning utilitarianism abstracting from the separateness of persons, it also expands the contrast set out in the previous section between utilitarianism and justice as fairness.

The first contrast Rawls draws in the sixth section is between the two theories' response to common sense morality. Again, as was pointed out in the previous posting, Rawls has Sidgwick implicitly in mind here given that the Methods of Ethics took seriously from the first the notion of "common sense morality". Naturally Rawls' response to the challenge of common sense morality differs from that of Sidgwick since Rawls assumes that justice as fairness can take the claims of common sense morality more seriously than can utilitarianism. The main reason given for taking this to be true is that the convictions of common sense morality concerning justice are assumed to be consequences of the principles that would be chosen in the original position. This assumption is quite different from the utilitarian one which adopts the more empiricist argument of viewing common sense morality as having a generic good function even though the rules of it are secondary to those of utilitarianism. (Rawls does not here invoke the general "Government House" conception that Sidgwick adheres to of taking the dictates of common sense morality to be ones that might be publicly adhered to whilst secretly one is committed to utilitarianism. Given Rawls' own conviction of the importance of the publicity of principles of justice and the way this conception cuts directly against requirements of common sense morality we have here a stronger argument for viewing justice as fairness as taking common sense morality more seriously than utilitarianism than Rawls gives in the text!)

The second contrast between justice as fairness and utilitarianism is the more familiar one that suggests the contractarian nature of the former is effectively more social in form than the principle of the latter given that the latter extends the principle of choice for one man to the whole of society. This contrast invokes the charge of abstraction from the separateness of persons against utilitarianism and so is simply a restatement of the argument of the fifth section.

The third contrast invokes the distinction between deontological and teleological approaches to ethics, a contrast related to the question of the priority of the right over the good or vice versa. The sense in which justice as fairness is a deontological view is specified here as being that it does not interpret the right as maximising the good. Whilst this point is familiar in some respects the means by which Rawls defends it in this sixth section is unexpected. It concerns the manner in which the good is taken to be prior to the right by utilitarians. If the utilitarian theory is viewed as giving priority to the good over the right the natural question arises concerning what the good is taken to be by this theory. Here Rawls makes the key point that the good is simply assumed to be that which is desired. In other terms, whatever is desired would appear to have value so the question of maximisation arises due to a prior commitment to this conflation of the good with the desirable. It is natural to wish to maximise the desirable and if the good is viewed through the prism of the desirable then the maximisation strategy is intuitively plausible as a whole.

However, following Kant's argument concerning the nature of happiness, Rawls does not accept a prima facie value as attached to whatever is desired. Rather, Rawls takes it to be the case that only that which matches the principles of right has value. The principle of right restricts (and thereby realises) what can be taken to be good. This is the point of the notion of the priority of the right over the good: the right provides us with formal characteristics that have to be satisfied in order that something can be taken to be good. In lieu of these being satisfied you merely have something that some may wish for but not something that has any status of value as such. As Rawls puts this: "desires and aspirations are restricted from the outset by the principles of justice which specify the boundaries that men's systems of ends must respect". So only certain kinds of ends will be taken seriously as only these ends are worthy of being taken seriously (just as Kant writes that it is only under certain conditions that one is "worthy of" happiness).

Because of the way the contrast between utilitarianism and justice as fairness has here been characterised the priority of the right over the good becomes clarified as central to the conception of justice that justice as fairness elaborates. It has a moral ideal that guides the selection of any possible end as being acceptable as a value and there is therefore no value in a basic sense prior to the formal criteria of right. This second set of contrasts hence makes much clearer the core conceptions of justice as fairness.

Justice As Reciprocity (II)

In the last posting I did on this topic the focus on methodological devices was prominent. Rawls reaches the conclusion, however, that there is a close relationship between the principles of justice and the sense of justice. And from this connection he states some key points, including the argument that justice is "the first moral virtue" as it arises from the concept of morality being related to self-interested persons in similar situations. Effectively, if it is possible at all to transcend the bounds of rational self-interest then this is the concept that we reach when we do so.


From the "conjectural derivation" as Rawls here terms the "original position" we have an assured root of both justice and fairness in reciprocity. Reciprocity is connected to the notion of free peoples relating to each other in a situation where no one possesses moral authority over another and we find ourselves engaged in some kind of mutual endeavour. Interestingly connected to this stress on reciprocity is the sense that just principles are ones that require people to be able to affirm them when they openly face one another. So Rawls builds in a publicity constraint on the principles that can be understood as ones of justice.


This publicity constraint is one that enables communicative relations between people and without it there can only be relations between people based on force. After building in this publicity constraint Rawls distinguishes fairness and justice from each other in order to be able to connect them back together. Fairness, on Rawls' view, applies to practices where there is either a competition between people or a cooperative activity in which they have no compulsion to participate. This is why the notion of fairness applies particularly to games and yet also to trade and bargaining. Justice, by contrast, applies to practices in which there is no choice whether or not one participates. There is, that is, an element of necessity in justice which applies to institutions that are generally pervasive or to practices which, if limited in application, are nonetheless ones that have no optional character for those involved in them.


Judging practices by reference to mutual acknowledgement of principles is to engage in applying the principle of reciprocity to their justification. So if one is engaged in a practice that we can acknowledge as satisfying the principle of reciprocity and so we can accept that its rules are either fair or just then we have no grounds of complaint with regard to it. So, for example, if the rules in question limit our sphere of activity in some important respects, then we have a right to expect that others will acknowledge in their own turn the same kinds of limits. If the practice is understood to be just or fair then there is a basic sense in which we all benefit from participation in it.


When one has thus acknowledged the practice in question as satisfying in this way the principle of reciprocity then it becomes the case that one has a "prima facie" obligation to abide by its rules. This obligation is what Rawls earlier referred to as the duty of fair play. Rawls admits, however, that in terming political obligation the duty of fair play he is extending the use of "fair" here beyond ordinary usage which only relates to acting in accordance with the spirit of a practice. But Rawls regards it as a not "unnatural extension" to incorporate the general obligation to abide by the rules in question as part of the duty of fair play as it requires that expecting the benefits of a practice to accrue to one does mean agreeing to the basis of that practice.


This point involves familiar considerations of "free-riding" that arose already in Rawls' earlier paper on justice as fairness. The interesting point about listing fair play as a prima facie duty is that in so doing Rawls deliberately adopts the language of W.D. Ross' intuitionism (though without referring to Ross). The justification of the general duty of fair play is clear in relation to the earlier mentioned publicity constraint since it meshes with this constraint in a natural way. Rawls further indicates that the recognition of the duty of fair play is not only closely meshed with the publicity constraint but that it is also an essential part of the recognition of another as a person. "To recognize another as a person one must respond to him and act towards him in certain ways; and these ways are intimately connected with the various prima facie duties." In making this point Rawls explicitly attempts to take the notion of prima facie duties away from intuitionist views and incorporate it instead in a basic act of recognition.


This point about recognition is used further in relating fairness to justice in terms of the earlier distinction between voluntary and involuntary engagement of practices. If fairness applies to practices in which one engages voluntarily then it is obvious that no one willingly agrees to play in a game that is stacked against them. So in relating the notion of fair practices to justice we effectively act as if practices that were carried out necessarily had in them the freedom we express in situations of choice. 


Having reached this point in the analysis Rawls turns next to a comparison between this notion of justice as reciprocity and the utilitarian view. The utilitarian view is taken to be in clear conflict with the notion of justice Rawls has advanced due to the over-riding concern of utilitarians being with efficiency in which administrators are concerned with rules that have little relation to justice. Utilitarian calculation is concerned with preferences and interests so that it engages in abstraction from the separation of persons. 


Rawls presses the case about the difference by considering the case of slavery. Utilitarian calculations concerning it have to begin with the interests of all concerned and, whilst they may well produce from this analysis, a condemnation of slavery, there is a fault at the root of such a conception as it does not relate to the slave-holder and slave in terms of reciprocity. If one has reciprocity at the root then you cannot regard the slave-holder's interests as in way relevant in making out the injustice of the institution that he perpetrates. This does not mean that it could never be treated as provisionally excusable but it could only be related to as such if there was an advance in the reciprocal understanding involved in it on what had previously been established and not at all with regard to interests of a slave-holder. 


The principles of justice that Rawls has specified could never select slavery as a bearable system and these principles have special weight in determining the character of reciprocity. Following on the earlier stress on the notion of recognition Rawls subsequently points out that all the elements of justice as reciprocity are involved in Kant's formula of humanity as an end in itself. The general point of the argument, however, and this is importantly different from where Rawls ends up in Political Liberalism, is to show the moral roots of the conception of justice that is being justified.

Rawls' Original "Original Position"

The device of the "original position" in Rawls' A Theory of Justice is one of the most famous elements of the work but there is little discussion of how he arrived at the idea. In his 1963 paper "Constitutional Liberty and the Concept of Justice", a paper republished in the Collected Papers, we find the original version of the notion of the "original position" minus the "veil of ignorance" which is not introduced for another 4 years. The description of the idea in this early paper is interesting both in its own right and as part of an account of the genesis of the developed notion in A Theory of Justice. 


This paper opens with a description of the notion of justice, a notion carefully distinguished from a general vision of a good society as it is "only one part of any such conception". What justice requires, however, is the elimination of arbitrary distinctions and the establishment of a "proper balance or equilibrium" between competing claims.  Naturally these points leave everything open as Rawls himself admits since it is only the specific conception of justice developed that will tell us what distinctions are arbitrary.


The conception of justice that is developed in the piece is intended to apply to political institutions which are described as "publicly recognized systems of rules", rules that are normally acted on and which define offices, rights and privileges. This institutional conception of justice is already pointing towards though it does not yet specifically state the notion of the "basic structure". 


As the title of this piece suggests Rawls was specifically concerned in it to discuss constitutional liberties although he includes here amongst such the notion of equality of opportunity (perhaps in response to the point about arbitrary distinctions). These liberties "define and establish an initial position of equal liberty" for all citizens in the social system.


Having defined this territory Rawls goes on to provide a non-utilitarian argument for these constitutional liberties and this argument draws on the notion of "justice as fairness" that was first elaborated in an article of 1958 which I discussed a little while ago. The notion is defined here in the following way:

The concept of justice which I shall use may be stated, for the moment, in the form of two principles: first, each person participating in an institution or affected by it has equal right to the most extensive liberty compatible with a like liberty for all; and, second, inequalities as defined by the institutional structure or fostered by it are arbitrary unless it is reasonable to expect that they will work out to everyone's advantage and provided that the positions and offices to which they attach or from which they may be gained are open to all. These principles express the concept of justice as relating three ideas: liberty, equality, and reward for services contributing to the common advantage. (75)

It is clear that the first principle has a relationship to the liberty principle in A Theory of Justice and that the second has a connection with some popular formulations of the difference principle. The key point about the second principle given here, however, is, again, its specific statement of a view of "openness to talents". It is intriguing that the third idea that Rawls here states as included in the concept of justice as fairness is one that is akin to the notion of social utility even though it is precisely to provide an alternative to this notion that Rawls is presenting his conception.


The principle of liberty is indicated to be surprisingly restricted since although it throws the burden of proof on anyone who wishes to restrict liberty there can be "and often is" a justification for doing so states Rawls. In commenting on the second principle it is also indicated that inequality is understood here not to affect differences within a hierarchy but rather "differences in the benefits and burdens" that arise from them.  The principles are applied to institutions and, in this application, Rawls introduces the methodological idea of the "representative man" and what he might be thought to hold. "Unless one is prepared to criticize the system of institutions from the standpoint of a representative man holding some particular office, one has no complaint against it." (76) Institutions can be said to be just or fair when those who participate in them could propose its rules for adoption in a situation of equal liberty.


It is after indicating that this is the way to understand the justice of institutions that Rawls introduces his original conception of the "original position". In this original conception it is framed as an "analytic construction". Within this construction we take a society of persons who are "normally self-interested", rational (with some conditions of this notion spelled out) and have sufficiently similar needs for interests to be broadly complementary. They already belong to a society of institutions so we are not envisaging them setting these institutions up. Rather, the device is here used to assess how they could determine what kinds of complaint against practices should be accepted as legitimate.


This procedure requires the persons in question to propose principles on which complaints can be tried, complaints that are subjected to three conditions. The conditions are: firstly, that such principles as are accepted will apply to complaints of others, secondly, that no one's complaints will be heard until everyone is of basically the same mind about the means by which they will be judged and, thirdly, that the principles proposed on any one occasion are binding for the future unless there are special circumstances. So, effectively, this process of adjudication of complaints creates a set of norms that provides the basis for cases to be aggregated together under general headings. The general point of this procedure is then highlighted: "The restrictions which would so arise may be thought of as those a person would keep in mind if he were designing a social system in which his enemy were to assign him his place" (78). Even if the most ill-willed person were to be taken as giving you your place in the social system you would be safe, as it were, given these guarantees of principle and process.


The suggestion is that this "analytic construction" would lead the persons in question to adopt Rawls' earlier cited two principles of equal liberty and an openness to talents that is meritocratic. Inequalities are thus embraced by those considering the "analytic construction" since, it is suggested, they would work as "incentives" to better efforts, incentives that can be considered as "concessions to human nature" (79).


The "analytic construction" is not, however, taken to be conclusive by Rawls as mandating his two principles as he considers what he terms a "natural objection" to it. The "natural objection" is that some might prefer to contract into a caste system rather than a society governed by the two principles of justice. Rawls argues against this objection both by considering the notion that persons know their talents and that they do not. If they do not know their talents then they need to consider the simple remoteness of the possibility that they could be in the highest caste. If talents are known, by contrast, the more able can point to the second principle as a guarantee of welfare for the less able.


The general argument of the piece also reaches a Kantian point towards the conclusion as Rawls indicates here that there is a peculiar feature of the concept of justice, namely, that it treats each person as an equal sovereign and "requires a unanimous acknowledgement from a certain original position of equal liberty" (94). The argument of the piece is, compared to the elaboration of the original position later, fairly primitive and the discussion of the counter-example, whilst anticipating Rawls' later problems with perfectionist accounts of society, is far from really discussing the assumptions involved in social orders that are not based on openness to talents. Despite these problems it is illuminating to see that the original "original position" is one that works within the constraints of an existing order and has the purpose of providing a rationale for complaints within existent institutions. So it was not originally conceived as a direct replacement for the social contract but more for a response to the utilitarian demand for justification of legitimate complaints.