The general preface to Chapter 2 makes apparent that there are, on Rawls' view, two parts to the theory of justice. The first part concerns an interpretation of the initial situation (the original position) which will be the subject of Chapter 3 and the second part concerns an argument establishing which principles would be adopted in this situation. Oddly Rawls treats the latter before the former. This is odd since the full argument for adopting the principles evidently cannot be given prior to the discussion of the original position so the formulation of the principles of justice in this chapter must be preliminary and yet it is still thought worthwhile, for reasons as yet unclear, to treat them before the formal constraints of the original position are introduced. The point of Chapter 2 as a whole is also said to be to "explain the meaning and application" of the principles, two presumably distinct tasks.
After this introductory paragraph of the whole chapter section 10 opens which is the first full section of Chapter 2. Rawls opens this chapter with his general statement that the "primary subject of the principles of social justice is the basic structure of society", a statement followed by an interpretation of the meaning of "basic structure" as consisting in the "arrangement of major social institutions into one scheme of cooperation". So if the "basic structure" consists of the "major social institutions" it is not surprising to find that Rawls does need to discuss the meaning of the term "institution". The "rights and duties" applicable to these institutions are taken to determine the general appropriate distribution of "benefits and burdens" in social life as a whole. But the principles which apply to institutions are also distinct from those that relate to individuals (as becomes further apparent in the final two sections of Chapter 2 where Rawls looks, after the principal business of the chapter has been concluded, at principles for individuals).
After making these general points distinguishing principles for institutions from principles for individuals on the one hand and defining the primary subject of justice in relation to major institutions on the other Rawls turns to one of the two primary points of section 10, namely the provision of what turns into a two-fold discussion of the "theory" of institutions. The first type of "theory" emerges as Rawls describes and defines the nature of what he means by an "institution". An institution is, he says, "a public system of rules which defines offices and positions" along with attendant rights, duties and so on. The rules in question distinguish between what is allowed and what is forbidden and indicate penalties if the rules are breached. At this point Rawls is describing institutions in a very general way as inclusive of such things as rituals. However, the elaboration of the account of institutions next goes in two distinct directions.
The first direction concerns the distinction between institutions considered abstractly and institutions concerned concretely. An abstract conception of institutions relates only to "possible" forms of conduct and thought whilst a concrete one deals with the realization of the institution. Although it is possible to describe both abstract and concrete forms of "institution" as just or unjust it is, in ordinary discussion, the concrete forms that attract most attention though the important point here is that if an abstract notion of an institution is unjust then this would vitiate any possible attempt to realize it, not just some specific way in which it had been realized. The understanding of concrete institutions is illustrated by Rawls' example of parliaments which are governed by certain rules in which general norms are meant to cohere together in an overall scheme. This example is specifically used by analogy with H.L.A. Hart's description of a legal system.
This first division of the subject is only of serious interest in that Rawls describes the general usage of just or unjust as applied to concrete institutions as the "best" use of the terms in relation to institutions, an odd view given that it is basically abstract institutions that we would be expecting from his own theory. More importantly than this peculiar point, however, is that the general description of an institution as a public system of rules allows Rawls, for the first time, to really say something about publicity. The sense that an institution is such a public system leads Rawls to assimilate the general notion of an institution to that of a contract. This is despite the fact that Rawls is aware that it is a "simplifying assumption" in relation to actual institutions (and it is to the latter that appellations of just or unjust would be "best" applied).
It is through the publicity of the rules in question that a general epistemic result is achieved: that of ensuring the norms expected of those engaged in its running are known to those so engaged. Now, if the institution in question belonged also to a "well-ordered society" (was, in this sense, an "abstract" one) then there would follow from this epistemic result "a public understanding of what is just and unjust". This is why when the following chapter deals with the constraints of right one of these is a principle of publicity and Rawls thus ties publicity and justice very closely together (at least in ideal theory).
Having made the distinction between abstract and concrete institutions Rawls has effectively made clear the basis for the former's theory of institutions to reside in a clear commitment to publicity and in this sense the discussion here parallels the Kantian account of publicity in Perpetual Peace. Having done this, however, the second sense of a "theory" of institutions comes into view in the section. Rawls borrows from John Searle a distinction between "constitutive rules" and "strategies and maxims". According to this distinction the institution is established by means of rules that broadly define rights and duties but the strategies and maxims relate to how the institution is used by those who belong to it. The latter are not themselves "part" of the institution but are said instead to be part of its "theory". In saying the strategies and maxims are part of the "theory" of the institution Rawls is using the term "theory" in a specific way and in a way that is distinct from the way we might think of the discussion of institutions up until this point in section 10 as indicative of a "theory" of institutions.
If the distinction between abstract and concrete institutions and the general definition of an institution as a "public system of rules" is indicative of a kind of "theory" of institutions then it is not a "theory" that is at work primarily in relation to how the members of an institution work within it. It is rather what one might call a "theoretician's theory" of the institution whilst the strategies and maxims adopted within the institution are, by contrast, indicative of a "practitioners theory". The "practitioners" have a theory of the institution in the sense that they want to work out how best to live and thrive within it, a "theory" that we might think of as a way of governing one's prudential and moral conduct within the institution's orbit. Such practitioners are not interested in the "constitutive" rules of the institution themselves except in relation to utilising them for their own ends (which may or may not be the "ends" of the institution itself). Further the general rules of the institution may encourage behaviours in its members that are far from being those which it was intended the institution should further. This is indicative of one of the problems of social reform since the ways in which practitioners' theories will develop would have to be taken careful account of by anyone who wished to alter the constitutive rules of an institution. As Rawls puts this: "The conduct of individuals guided by their rational plans should be coordinated as far as possible to achieve results which although not intended or perhaps even foreseen by them are nevertheless the best ones from the standpoint of social justice". So the point of the way the institution is set up would be to produce an end much like the one Adam Smith attributed to the market as one in which the rational and prudential aims of individuals tended to converge with what was socially best on the whole without it being necessary that the end of producing such a result be itself required to be adopted by anyone in particular. This "hidden hand" would be, as it were, built into the institution or such is the aim of social reform to help bring this about.
Having set out first the distinction between abstract and concrete institutions and then having made the distinction between two types of theory of institutions Rawls next looks at the different ways institutions could be unjust. The point made here is that institutions are not necessarily in themselves unjust as such just because some of the rules by which they operate are. It is also possible, within a social system, that the injustices of some institutions are outweighed by the virtues of others or compensated in some way by the latter. Finally a social system could be unjust even though none of the specific institutions making it up was as a result of something in the combination of these parts together.
Having made these general remarks about the different possible scopes of application of the notion of just or unjust in relation to institutions Rawls mentions, but only to discard, a possible meaning of a "theory" of justice as a way of articulating again his own conception of a "theory" of it. Rituals are, as Rawls puts it, not usually regarded as being in themselves just although they clearly are types of institution, so would belong within what he terms a "general" theory of institutions. But this "general" theory which might be part of a very general account of "justice" is not what Rawls wants to present as he is only concerned with the "basic structure" of society and therefore with what he terms "the standard cases of social justice".
So the "theory" of justice is a "theory" of social justice, a type of theory that emerges from a restricted (and not a "general") theory of institutions. In this "theory" the emphasis is on the "basic structure" and its rules. The "basic structure", to emphasize again, concerns the "fundamental rights and duties" and therefore the "division of advantages from social cooperation". Now, given this account, any social system one can imagine will be governed by the rules of its "basic structure" and hence will enshrine a certain type of conception of justice even if the general conception in question is not manifestly visible at all times or to all peoples (since the society may not be that well-ordered and so may not adhere very clearly or consistently to a principle of publicity).
Within the general conception of any given basic structure it is to be expected that there could, even within a set-up that was in many ways "unjust", be a commitment to impartial application of the rules by which institutions are governed. If a commitment to such impartial administration did exist then the institutions would therefore treat similar cases in a similar way. Such a view of justice, based on the impartial and consistent application of the norms in question, is presented by Rawls as a "formal" conception of it. "If we think of justice as always expressing a kind of equality, then formal justice requires that in their administration laws and institutions should apply equally (that is, in the same way) to those belonging to the classes defined by them."
This formal conception of justice that Rawls here entertains is taken, as we discovered the notion of "classical utilitarianism" was, from Sidgwick. Such a formal notion of justice is also described by Rawls as "adherence to principle" or "obedience to system" (and these characterisations perhaps have ominous echoes of the "Eichmann defence" from his infamous trial). This conception is clearly insufficient to prevent injustice since discrimination between groups, even to the point of slavery, is quite consistent with it. Rawls also describes formal justice as "justice as regularity". Although it is clearly a conception that appears problematic it is not true that the notion of it does not have important things to it that are worth consideration since the commitment to impartiality embedded in its application rules out corruption of officials. Consistency of application further ensures that even unjust rules create stable patterns of expectations which prevents arbitrary rule.
This discussion of formal justice is the final subject of section 10 and it ends with an inconclusive contrast being drawn between formal and substantive notions of justice. The reason why this contrast emerges is due to the allegation (made by Lon Fuller amongst others) that a purely formal system of justice could never (or very rarely) be administrated impartially. The nature of discriminatory institutions set up to serve some at the expense of others is such that the practitioners theory of those running such institutions would be one that permitted them to alter constitutive rules as they saw fit and hence could not create stable patterns of expectations. Such an accusation is akin to Kant's remark, again in Perpetual Peace, that non-republican systems of government ("despotisms" as he termed them) are essentially without form. Holding such a view effectively leads to the conception that the "formal" conception of justice indicates something like a state of nature in operation.
Under the objection Fuller made to the "formal conception" of justice there is an indication of a need for law-makers and leaders of institutions to be required to commit to a conception of justice in order for the institutions to have any way of even meeting truly formal requirements as are expressed in the notion of the Rule of Law. On this kind of view it is only if there is a substantive notion of justice that legislators are tied to that the possibilities of impartial administration can be satisfied. However although Rawls mentions this view of Fuller's here (a view that we have seen as having a formal connection to Kant's view) he does not investigate its contentions at this point. Rather the section concludes with the argument that the Fuller view can only be assessed after the work of the rest of Chapter 2 and the whole of Chapter 3 has been done. So Rawls here sets down a promissory note indicating that the Fuller view will be returned to and assessed but only after the real subject of justice in Rawls' own view has been suitably addressed in sufficient detail.