The shift announced from a concern with the "good" to that of "stability" is one that hides the continuity between Chapters VII and VIII. Just as Chapter VII concluded with an account of the good for persons that included a rudimentary theory of virtue and vice, so also the first few sections of Chapter VIII concerns different levels of moral development, including thereby a description of moral sentiments. Indeed, the basic subject of Chapter VIII is nothing else than a general account of moral psychology. Given that this is so, there is rather less of a shift in focus from Chapter VII to Chapter VIII than Rawls' introductory remarks to the latter suggest. In fact, within Chapter VII it was stated that it concerned the good for persons and that later Rawls would look at social goods. This does mark the difference between the discussion in Chapter VII and that in Chapter IX. Chapter VIII presents, by contrast, an intermediate level of consideration which is meant to show how the good for persons is best viewed as embedded within a social sense of the good.
If Rawls' contrast between the foci of Chapters VII and VIII is, to an extent, misleading, however, there is still some sense in viewing Chapter VIII as providing considerations of a sort that Chapter VII did not include. What is fundamentally at issue in Chapter VIII is a preparation for the concluding chapter in which the congruence of the sense of justice with the sense of our own good is laid out and this was only hinted at in Chapter VII whereas Chapter VIII consistently indicates a concern with this question.
The first section of Chapter VIII concerns the concept of a well-ordered society (WOS). This conception was first described at the very beginning of Theory where it was determined as a society "designed to advance the good of its members and effectively regulated by a public conception of justice", a determination repeated here. Having restated this notion Rawls next spends some time expounding not, initially at least, on the idea of the "good" of the members of the society but, instead, on the sense of a "public conception of justice" (my italics). The key point about this is that "everyone accepts and knows that the others accept the same principles of justice". This public point is subsequently stressed as essential to justice as fairness. This occurs through first a discussion of the original position and then an account of the WOS. The original position is framed in such a way that the principles chosen are assumed to be ones that can be publicly justified and this is part of the way that the probable effects of adopting principles of justice is assessed there. Any conceptions of justice that depend on esoteric elites holding back knowledge are rejected on principle and the conception of justice adopted is one that is assumed can be based on generally available knowledge concerning people and their place in society, a consideration that has importance in subsequent sections of Chapter VIII.
If the original position is thus constrained by reference to publicity conditions it follows that the conception of the WOS will have to fulfil the conditions that were specified during the course of deliberation in the original position. Hence the members of the WOS will have the desire to act in accordance with the principles of justice given that these principles will be known to regulate the conduct of all within it. It is after specifying the sense of public adherence to the principles of justice as a central feature of the WOS that Rawls turns, for the first time in this chapter, to considering the idea of stability. A conception of justice is more stable if "the sense of justice that it tends to generate is stronger and more likely to override disruptive inclinations and if the institutions it allows foster weaker impulses and inclinations to act unjustly". In presenting the test for the stability of conceptions of justice in this way Rawls presents what are, effectively, two different tests. On the one hand a conception of justice is more stable if it has a resilient psychological appeal whilst on the other hand it is stable if, by following it, we are led to construct institutions that, in their elementary functioning, discourage unjust inclinations. The first of these tests of stability is one that can be assessed by means of moral psychology, the second, by contrast, requires a conception of social institutions. This marks the real difference between Chapters VIII and IX as Chapter VIII addresses the stability test in relation to moral psychology, whilst the discussion of congruence in Chapter IX, by contrast, is meant to outline the way in which the good of persons can be connected to a view of social institutions.
The stability tests thus require to be built into the original position just as it was regulated by requirements of publicity. The need for such a test in relation to moral psychology is evident in the sense that, without such a test having been ventured, there are no grounds for considering the anthropological realism of the conception offered. So the concluding chapters of Theory are meant to provide a basis for the claim that justice as fairness is a more stable conception of justice than others on offer. This is so despite Rawls admitting here that the criterion of stability is not alone decisive. What is meant in stating this is that it is possible for a view of justice to be advanced which takes little notice of the criterion of stability and Rawls interprets Bentham's utilitarianism in this way. However, even should this interpretation of Bentham be correct what it would show would only be that such a doctrine was a limit case with regard to conceptions of justice since the majority of such conceptions adjust themselves in more or less explicit ways to some features of what is taken to be humanly sustainable. The stability criterion is effectively meant to show how a conception of justice would in practice generate its own support and in Chapter VIII this test is applied by virtue of a general theory of moral sentiments being provided.
Within the first section of Chapter VIII Rawls describes the notion of stability as part of a theory of systems, systems that have reached equilibrium. In so doing Rawls draws upon some views developed within economics that are specified in the following way:
Three things are essential: first, to identify the system and to distinguish between internal and external forces; second, to define the states of the system, a state being a certain configuration of its determining characteristics; and third, to specify the laws connecting the states.
On this conception a system is in stable equilibrium when departures from it call into play forces that will tend to bring it back to its initial state. This function will be general assuming the shocks are within reasonable boundaries. Equilibriums are unstable whenever shocks force great changes within the internal operation of the system. The ability to adjust well to such shocks should be manifested within a time frame that is one that those within the system are able to bear though this notion is naturally left vague.
As far as Rawls is concerned the system in question is the basic structure of the society, a "complex", as he terms it, of political, economic and social institutions. In assessing the relative stability of conceptions of these the operative assumption throughout Theory is that of a generally self-contained national community, which is clearly a simplifying assumption. It does not imply that there is no change in the institutions but that any change within them is governed by continuous reference back to the conceptions of justice that is meant to govern their operation. One of the things taken to assure this is the development of moral sentiments that support the conceptions of justice.
Having arrived at this picture Rawls pauses to consider theories of moral sentiments and refers to two general traditions with regard to them. On the one hand there is the empiricist theory that has guided utilitarianism and which Rawls takes to also be reflected in the form of social psychology known as "social learning theory". A major contention of this theory is that the point of moral training is to supply what are termed "missing motives". What is taken to be missing is any original disposition towards principled right action. Given that this is missing it is the task of society to encourage such dispositions artificially. This is done by use of authority to mark approval and disapproval of conduct and to do this by means of rewards and punishments in order to produce a general sense of right and wrong. This view can be seen to be a kind of Pavlovian social conditioning view of moral sentiments. It is backed up by the sense that it is necessary to acquire moral sentiments at a stage prior to being able to understand them. Rawls thus views psychoanalytic conceptions of social learning to be a variant on the general empiricist tradition.
By contrast to this tradition the other view, which Rawls terms "rationalist" is associated by him with Rousseau, Kant, Mill and Piaget in which the development of innate capacities is encouraged. Taken in the generic sense in which Rawls pictures it this latter tradition is viewed as assuming natural sympathy exists between persons which provides an affective basis for the moral sentiments. Mill hence speaks of acceptable principles of reciprocity and tendencies to sociality. Such a model does not primarily stress external authority for social norms or the acquisition of new motives but rather the development of capacities already present towards their appropriate maturation.
Whilst it might have been expected that Rawls would view the latter position as more congenial with justice as fairness than the former he rather states that he assumes that there is much sound in both traditions and that he will try to combine them in what he terms a "natural" way. The subsequent next three sections of Theory aim to do this through an ideal picture of moral development within a WOS.