2. Two Senses of “General”

The phrase “a general conception of justice” requires clarification. Two senses of the phrase can be found in Rawls’s work. The first sense is a conception of justice that is general rather than special. In A Theory of Justice, he describes the following as a general conception of justice: “All social values — liberty and opportunity, income and wealth, and the social bases of self-respect — are to be distributed equally unless an unequal distribution of any, or all, of these values is to everyone’s advantage” or, for short, “Injustice […] is simply inequalities that are not to the benefit of all” (Rawls, Reference Rawls1999a, p. 54). He describes this conception as being “extremely vague and requir[ing] interpretation,” which his special conception of justice — namely, the two principles — aims to provide (Rawls, Reference Rawls1999a, p. 54). In this sense, then, a general and a special conception of justice are about the same topic, the latter being a (preferred) way of more clearly and precisely interpreting the former.

The second sense is a conception of justice that is general rather than about only one type of justice. In Justice as Fairness: A Restatement, Rawls clarifies that “Justice as fairness is a political, not a general, conception of justice,” because its “focus is almost entirely on the basic structure as the subject of political and social justice” (Rawls, Reference Rawls and Kelly2001, pp. 10–11). Rawls sees this focus as being only one of three “levels” of justice: “local” justice, “domestic” justice (a term he uses interchangeably with “political” justice and “social” justice), and “global” justice (Rawls, Reference Rawls and Kelly2001, p. 11). Rawls’s two principles are domestic-not-general because they are about only domestic justice rather than all types of justice. In this second sense, then, a general conception of justice would concern what is common to all types of justice.

Comparing the two senses, we see that what Rawls calls “a general conception of justice” in Theory, which is general in the first general-not-special sense, is not general in the second general-not-only-about-one-type sense. Theory’s general conception of justice is not plausibly a statement about what is common to, for example, both domestic and global justice, since Rawls’s account of global justice in The Law of Peoples does not require the elimination of all inequalities in social primary goods that are not to the benefit of all (Rawls, Reference Rawls1999b). Rawls does not offer (and, as already mentioned, opposes the idea of offering) a conception of justice that is general is the general-not-only-about-one-type sense. To avoid the confusion of these two senses, I will reserve the phrase “a general conception of justice” for a conception of justice that is general in the general-not-only-about-one-type sense, and I will use the phrase “a general conception of domestic justice” to refer to a conception of domestic justice that is general in the general-not-special sense. My interest lies in the general-not-only-about-one-type sense. I will leave aside whether my general conception of justice is also general in the other sense.

One extra wrinkle is that there are also two senses of the phrase “a political conception of justice” found in Rawls’s work. Indeed, when he writes that justice as fairness is political, not general, this sense of political is the less common of the two, the more common being political-not-metaphysical — that is, a conception of justice that is not based on a particular comprehensive doctrine (Rawls, Reference Rawls1996). Justice as fairness is thus a political (not metaphysical) special (not general) conception of domestic justice. My interest is in a general (not only about one type) conception of justice, and I will leave aside the question of whether my general conception is political-not-metaphysical.

3. Rawls’s Three (or Four) Parts of Justice

As mentioned, Rawls sees justice as having three parts, or levels: local justice, domestic justice (which he also calls “political” and “social”), and global justice. I have doubts about whether this is the best typology of justice (that is, a view about what types of justice there are and how they relate), but it should be laid out. Regarding global justice, by which Rawls means “principles applying to international law,” his theory “starts with domestic justice” and then “works outward to the law of peoples” (Rawls, Reference Rawls and Kelly2001, p. 11). Justice as fairness “applies first to the basic structure and sees [the question of] global justice […] as calling for separate consideration on [its] merits” (Rawls, Reference Rawls and Kelly2001, p. 11). Regarding local justice, the principles of political justice “constrain (or limit), but do not determine uniquely, the suitable principles of local justice” (Rawls, Reference Rawls and Kelly2001, pp. 11–12.) Separate principles of local justice are needed because “the basic structure and the associations and social forms within it are each governed by distinct principles in view of their different aims and purposes and their peculiar nature and special requirements” (Rawls, Reference Rawls and Kelly2001, p. 11). What he means by “social forms” becomes clear later: “[T]here are appropriate conceptions of justice that apply directly to most if not all associations and groups, as well as to the various kinds of relationships among individuals” (Rawls, Reference Rawls and Kelly2001, p. 164; see also Rawls, Reference Rawls1997, pp. 788–789). Those relationships include personal and familial ones; and so, “local” justice, as Rawls uses the term, includes justice within personal relationships, within families, and within associations.

There is another part of Rawls’s writing that could be interpreted as a fourth part of justice: his basic needs principle. He notes that the liberty principle “may be preceded by a lexically prior principle requiring that basic needs be met” (Rawls, Reference Rawls and Kelly2001, p. 44; see also Rawls, Reference Rawls1996, p. 7). He does not write much more about this than the following:

There is a question of whether the basic needs principle should be viewed as the first piece in Rawls’s conception of domestic justice, or, instead, as a distinct part of justice. That question of Rawlsian interpretation is not crucial for this paper’s goals, but I will briefly argue for viewing his basic needs principle as a prior part of justice.

Recall his general conception of domestic justice: in short, “Injustice […] is simply inequalities that are not to the benefit of all” (Rawls, Reference Rawls1999a, p. 54). This suggests that domestic justice, as Rawls sees it, is egalitarian in its entirety, in the sense that it is concerned only with distinguishing between just and unjust inequalities, which it does by seeking to determine which inequalities are to the benefit of all. His two principles, as a special case of that general conception, are all about specifying which inequalities are just and unjust, which we can see by noting that equal basic liberties, fair value of political liberties, and fair equality of opportunity describe how inequalities of certain social primary goods are not (arguably) to the benefit of all, and the difference principle describes how inequality of other certain social primary goods is (arguably) to the benefit of all. In contrast, a basic needs principle is sufficientarian rather than egalitarian, since it aims to identify an adequate “level of material and social well-being, and of training and education” (Rawls, Reference Rawls1996, p. 166). Answering that question of adequacy does not help to provide a clearer and more precise interpretation of Rawls’s thoroughly egalitarian general conception of domestic justice. The sufficientarian demand of the basic needs principle must be met before egalitarian questions can be salient, which suggests that his basic needs principle is about a distinct, and prior, part of justice.

Although I would not defend Rawls’s typology of justice, I wish to preserve three of his insights. The first is his explicit statement that there is more to justice than the sort of justice that his two principles seek to illuminate. Instead of using the language of “levels,” I will refer to this as the claim that there are multiple types of justice. Second, I follow Rawls when I claim that each type of justice has its own principles, because “of their different aims and purposes and their peculiar nature and special requirements” (Rawls, Reference Rawls and Kelly2001, p. 11). Third, a desideratum of my general conception of justice is that it should be able to make room for, and elucidate what is common among, the various types of justice Rawls mentions, including: a sufficientarian concern for basic needs, an egalitarian concern for fair terms of cooperation between productive cooperators, a concern for justice within associations since they are sustained by powerful social norms, and a concern for justice within personal and familial relationships, which are also impacted by powerful social norms. I have not included global justice as a separate item on this list, because I resist drawing so sharp a line as Rawls does between domestic and global justice, instead considering the extent to which any type of justice has (or can acquire) an international or global scope.

4. Justice as Environment-Shaping Responsibility

I will now describe my general conception of justice: justice as environment-shaping responsibility. In short, my view is that injustice, in its general form, is the misuse of environment-shaping power. An environment-shaping action is an action that changes the rules, norms, or physical features of an environment with the aim of making some things easier for at least some people. Environment-shaping power is the power to take such actions, which tends to be unequally distributed. One example of environment-shaping is how (re)designing the basic structure of a society changes the rules to make large-scale productive cooperation easier among its members. Justice is much more than the justice of a society’s basic structure because there are many other environment-shaping actions. For example, part of the breadth of the idea of environment-shaping can be seen from the social model of disability, the key message of which is that the degree to which, or whether, a person is disabled depends on how people’s earlier actions have shaped aspects of the person’s environment(s), such as the physical accessibility of built environments, the inclusiveness of social norms, and the prevalence of disability-positive attitudes. On models of disability, see, for example, David Wasserman and Sean Aas (Reference Wasserman, Aas, Zalta and Nodelman2023).

According to my general conception of justice, justice in its most general form demands that the people (or agents, which could be corporate) who have the power to shape an environment use their power in ways that are accountable to the people who are owed justification. For example, the design of a society’s basic structure affects, at least, all of its members who are potentially productive participants in the cooperative system, as well as perhaps some people outside of the society who are affected by the way that society’s basic structure enables and hinders certain forms of cross-border productive cooperation, and possibly others as well. The question of who the people are who are owed justification because of how they are affected by the environment-shaping power of others needs to be addressed for each type of justice.

I will now offer an argument for my general conception of justice as environment-shaping responsibility. The argument starts with an analysis of Rawls’s general conception of domestic justice and then shows how we can get from there to justice as environment-shaping responsibility through a process of generalizing. Recall Rawls’s general conception of domestic justice: “Injustice […] is simply inequalities that are not to the benefit of all” (Rawls, Reference Rawls1999a, p. 54). He has in mind inequalities in life prospects as these are affected by the design of the basic structure of a society — a design that is achieved by the use of the coercive power of a government. (I will say more below about two ways in which a government’s power is coercive.) Rawls’s general conception of domestic justice can be rephrased, then, as: Domestic injustice is the use of coercive government power to design a society’s basic structure that yields inequalities in life prospects that are not to the benefit of all. This phrase has three pieces: (i) the fact that coercive government power is used, (ii) what the power is used for — namely, to design a society’s basic structure, and (iii) the condition that makes such a use of power unacceptable — namely, that it yields inequalities that are not to the benefit of all. I will discuss each piece.

Regarding the first piece, in addition to coercive government power, the power of social norms is also relevant to justice, especially justice within associations and within personal and familial relationships. I will refer to the former as “political” power and the latter as “social” power. A key difference between these two types of power is the extent to which they are coercive. Political power is coercive in (at least) two ways: first, it is ultimately backed up by the government’s claim to a monopoly on the legitimate use of force: “Political power is always coercive power applied by the state and its apparatus of enforcement” (Rawls, Reference Rawls and Kelly2001, p. 40). Even if a member of a society ends up never being personally subject to state violence, their degree of compliance with government rules is affected by their knowing that excessive non-compliance may result in physically enforced penalties. Second, as Rawls argues,

In other words, people should have the right to emigrate (except where a restriction on that right is a justified punishment), and, of course, some people do voluntarily exercise that right, but reluctant political emigration — that is, wanting to stay, but leaving to escape a government’s authority — has personal costs so great that it is not a voluntary choice; or, more precisely, justifications of how political power is used should be made as if leaving is not an option.

In contrast, we might be tempted to say that social power — that is, the power of social norms — is not coercive in either of those two ways, since violence to enforce social norms is (or should be) illegitimate, and since freedom of conscience, freedom of association, and other civil liberties (should) adequately protect a person’s ability to voluntarily leave an association or a relationship or to otherwise make non-conforming choices. For example, Rawls argues that

But this is too quick. First, as Iris Marion Young notes, the danger of social power lies in part in its potential to normalize illegitimate norm-enforcing violence: “What makes violence a face of oppression is less the particular acts themselves, though these are often utterly horrible, than the social context surrounding them, which makes them possible and even acceptable” (Young, Reference Young1990, p. 61). Second, the personal costs of leaving an association can also be great, even when relevant legal rights are well protected. And third, although Rawls is correct that no association comprises all of society, some social norms — such as those concerning race, gender, sexuality, and other aspects of identity — are more encompassing and harder to escape, with sometimes significant rewards and penalties for conforming or not. I maintain that the distinction between political power and social power is still worthwhile, but the above considerations narrow the difference between them.

The second piece concerns what the power is used for — namely, to design a society’s basic structure. This is an environment-shaping action: using political power to (re)design the basic structure of a society changes the rules within the boundaries of that political authority in ways that aim to improve the extent to which the regulatory environment facilitates productive cooperation on a large scale — that is, among millions of people who work within that system of cooperation (where such work includes reproductive labour, parenting, and other care work, whether paid or not). There are, however, many other possible environment-shaping actions, including actions that change the regulatory environment to facilitate other aims (such as unconditionally meeting basic needs), and actions that change one or more other aspects of an environment (such as social norms, attitudes, or physical spaces). For example, consider efforts to change norms for rounds of introductions to include an invitation for people to give their pronouns along with their names, which is environment-shaping that aims to improve the trans-inclusiveness of social environments. Justice as environment-shaping responsibility can see this as an issue of associational or personal justice, where those with more power to shape the norms in question should be accountable (morally and socially, at least) to those most affected by the current shape of the relevant social environments.

The third piece concerns the condition that makes a specified use of environment-shaping power unacceptable. Once we acknowledge that injustice can come from the use of two kinds of power — political and social — to do a variety of environment-shaping actions, we should not expect that there is only one condition that can make any use of environment-shaping power unacceptable. And so, a general conception of justice should refer to environment-shaping power in general.

Based on that analysis, we can use a process of generalizing to move from my rewording of Rawls’s general (not special) conception of domestic justice to my general (not only about one type) conception of justice as environment-shaping responsibility. Recall my rewording of Rawls: Domestic injustice is the use of coercive government power to design a society’s basic structure that yields inequalities in life prospects that are not to the benefit of all. When we move from domestic justice to justice in general we want to (i) consider social power as well as political power, (ii) consider any environment-shaping action, not only those designing a society’s basic structure, and (iii) not specify one condition (such as yielding inequalities that are not to the benefit of all) as the only condition for the acceptability of any use of environment-shaping power. Those changes give us: Injustice is the use of political or social power to take environment-shaping actions that are unacceptable . A duty of justice is therefore a moral demand to be accountable for the way we use power to take environment-shaping actions. And so, injustice is the misuse of environment-shaping power, and justice is environment-shaping responsibility.

5. An Objection and Reply

Consider the familiar idea that if a person commits a serious wrongdoing (which need not involve them misusing their environment-shaping power), the wrongdoer ought to be “brought to justice,” so that they can get their “just deserts” — namely, the punishment they deserve. That idea is about purely retributive punishment, which is distinct from questions about how to improve the safety of environments (where safety should be understood to mean the safety of persons first, both from individual and state violence, and of possessions second). Justice as environment-shaping responsibility clearly includes questions of what should be done to improve the safety of environments, and so, it could include some questions of punishment if punitive measures (such as those focusing on deterrence or rehabilitation) should be seen as a necessary part of a society’s strategy to achieve a desirable degree of safety. (I leave open the possibility that alternatives to punitive measures are more defensible.) But let us imagine an objector who thinks that punishment is sometimes deserved when it serves no purpose other than retribution, and that this is a demand of justice. The objector might worry that my general conception of justice leaves out retributive justice.

The objector would be right. I do not see purely retributive punishment as a part of justice, even if it is a part of morality more broadly. In short, my view is that “just deserts” are just “deserts.” I believe that we can say whatever ought to be said about purely retributive punishment — whether to defend its occasional appropriateness or to argue for its abolition — by talking about desert without combining it with justice. The idea of purely retributive punishment relies on the moral question of what bad things a person is owed as result of their own morally culpable (in)actions — a question complicated by difficulties in drawing a sharp enough line between a person’s choices and their circumstances. We don’t need the term “justice” to discuss that question, since the term “desert” suffices.

Compare this with the moral question of what a person is owed as result of the environment-shaping other people have done or have the power to do — a question that, like any question of moral responsibility, is also complicated by difficulties in drawing a sharp line between the choices and circumstances of the people who have taken (or have the power to take) environment-shaping actions. (Since it affects both of those moral questions, I will not take a stand here about the choices-circumstances debate.) Consider some examples of this second moral question: those who are disadvantaged or harmed by heteronormativity are owed more inclusive norms, and people bear that duty in proportion to how much environment-shaping power they have to improve the relevant norms; people with mobility impairments are owed accessibility improvements to older buildings as a result of past designers’ choices about the built environment, as well as being owed accessible building codes going forward; and going back to Rawls, as a result of regulatory environment-shaping that aims to allow productive socioeconomic inequalities, the least advantaged in a society are owed principles regulating those inequalities that allow them to see the inequalities themselves as affirming, not challenging, their equality with more advantaged cooperators (which Rawls argues the difference principle achieves through the way it expresses reciprocity — see, e.g., Rawls, Reference Rawls and Kelly2001, pp. 126–130).

The questions of purely retributive punishment and questions of what I have been calling “justice” are about distinct reasons for what a person is owed. To say that a person who is morally culpable for a wrongdoing is owed purely retributive punishment is a different kind of moral claim than saying that a person who is negatively affected by the current shape(s) of their environments is owed a justification from those with the power to change it. Morality is by no means exhausted by those two kinds of moral reasons; surely there is more to morality than this. Here I am concerned with only the part of morality that is about what a person is owed as a result of how others have used and could use their environment-shaping power, and I suggest using the term “justice” to describe precisely that.

An alternative would be to distinguish between retributive justice and environment-shaping justice. That terminology is not unacceptable, yet to me it seems less clear. A perfectly good term already exists to help us consider questions of purely retributive punishment: desert. Using ideas of desert without combining them with justice does not reduce their importance. In contrast, we lack an alternative term for environment-shaping justice, and so, I suggest reserving the term “justice” for that idea. Of course, terminological preference is less important than the distinction itself, and so, I would not oppose those who would prefer to describe the distinction as being between retributive and environment-shaping justice — in which case, justice as environment-shaping responsibility would not be a fully general conception, since it would be described as being about what is common to types of environment-shaping justice.