A major topic in contemporary political philosophy concerns the characterization and defense of the ideal of public justification. This ideal holds, at a minimum, that the most fundamental political and legal institutions of a society should be publicly justified to its members. Esoteric justifications are excluded, as are justifications that proceed from premises that cannot be reasonably accepted. In the words of one of its proponents, the ideal demands that the political order “should in principle be capable of explaining itself at the tribunal of each person’s understanding.” Footnote 1 It is instructive to ask how the requirement of public justification differs from the ostensibly weaker requirement that the most fundamental political and legal institutions of a society should be justified (that is, supported by good reasons) and that the justifications should be done in public (as opposed to in private or not at all). The weaker requirement may be rejected by some. Perhaps postmodern skeptics will reject the whole project of political justification. And government house utilitarians may on occasion favor “noble lies” by the ruling elite. Still, the weaker requirement is not particularly controversial. By contrast, the ideal of public justification has generated persistent controversy and debate.
This essay presents an account of public justification that moves it considerably closer to the weaker requirement. The ideal of public justification, I argue, is, first and foremost, an ideal of rational justification. It is not an ideal that recommends a compromise with irrationality or rational incompetence. The account of public justification presented here aims to purify the ideal of elements that threaten to undermine its status as an ideal of rational justification. The result is a pure ideal of public justification, one that differs markedly from how the ideal is commonly characterized, but retains its distinctiveness as an ideal of politics. Footnote 2
I. The Moral Underpinning of Public Justification
The key feature of public justification is that it is relational in a way that (non)public political justification is not. This needs to be explained. But, first, some preliminary points are in order. Public justification is neither a maxim of political strategy nor a norm of political prudence. True, if we are to succeed in politics, then often we will need to justify our political proposals in public to a wide range of our fellow citizens. Likewise, in this or that circumstance, the public justification of the major political institutions of a society may be a condition for their stability. But these practical considerations do not capture the moral significance of the ideal. Footnote 3 We need to identify the moral considerations that underpin public justification before we elaborate its content; for these considerations must guide us in the elaboration.
It might be thought that no deeper moral considerations need to be identified. Public justification is an instance of moral justification. And if someone were to ask what explains the moral significance of having political arrangements Footnote 4 that are morally justified, it is not clear what could be said in response to him. But public justification is more than moral justification. It requires that political arrangements be morally justified in public — or perhaps, more weakly, that state officials, if called upon to do so, are able and ready to offer moral justifications in public for them. The moral significance of this requirement of publicity needs to be explained. Moreover, as I have said, to say that the political arrangements of a society should be publicly justified is not the same thing as saying that state officials should offer sound moral reasons in public that support them. Let me call this latter view the moral reasons view of political justification.
How might these two views of political justification — the moral reasons view and public justification — be related? With respect to a designated set of political arrangements, several possibilities can be distinguished. The arrangements may be condemned by both views (1), or they may be justifiable on both views (2). The institution of slavery, for example, is not justifiable on the moral reasons view nor could it be publicly justified to those who were subject to it. A bill of rights that safeguards certain fundamental liberties, by contrast, may be both publicly justifiable and justifiable on the moral reasons view. Two other possibilities are more interesting. An arrangement may be publicly justifiable, but not justifiable on the moral reasons view (3), or, conversely, an arrangement could be justifiable on the moral reasons view, but not publicly justifiable (4).
We will be interested in possibilities (2), (3), and (4). Possibility (1) poses no interesting question about the nature of political justification. Possibility (2) does, however. It is possible that a fully justified political arrangement is both publicly justifiable and justified on the moral reasons view. Likewise, possibilities (3) and (4) are plainly important, since they invite us to think about which of the two views is more fundamental to political justification. Highlighting these possibilities also helps us to see more clearly why the moral significance of the ideal of public justification cannot be explained simply by pointing to the value of having state officials offer good moral reasons in public in support of the political arrangements of the society over which they rule.
Public justification cannot be reduced to the view of political justification expressed by the moral reasons view because it purports to be relational in a way that the moral reasons view does not. Public justification involves justifying a particular political arrangement (or set of arrangements) to a particular person (or group of persons). Understanding this relational dimension is crucial to understanding the content of the ideal.
Return now to the issue of the moral considerations that underlie, or explain the significance of, the ideal of public justification. A tempting first thought is that public justification gives expression to the older ideal of government by consent. Footnote 5 Once government by consent is no longer seen to be a realistic possibility, public justification assumes the mantle of political voluntarism. This thought was at least suggested by Rawls. He claimed that a political society in which the political arrangements were publicly justified to all its members would come “as close as a society can to being a voluntary scheme.” Footnote 6 It seems fair to say, however, that proponents of public justification in general have not been too interested in seriously exploring proposals for making modern political societies more voluntary. Footnote 7 Rather than defending political voluntarism under modern conditions, they are better seen as proposing a departure from it. They seek to ground public justification not in respect for the will of persons, but in respect for the reason or the rational capacities of persons.
I too will follow this path. But the moral basis of public justification, on this latter understanding, will remain obscure until we have a better sense of what it means to respect a person’s reason or her rational capacities in the domain of politics. The bulk of this essay is an effort to fill in the details of this picture.
Start with a simple example. A government imposes a nighttime curfew and offers no justification at all for this course of action. Nor do its subjects have an opportunity to debate the matter before it does so. Now suppose that this government, in fact, is motivated by good moral reasons to impose the curfew, even though it never discloses these reasons to its subjects. In directing them to obey the curfew, the government does not engage their capacity to respond to reasons. Footnote 8 This capacity, of course, is not incidental to who these subjects are. Having this capacity is essential to being a person.
Respect for persons requires that governments treat their subjects in a manner that engages their capacity to respond to reasons. This requirement — call it respect for rational agency — explains the importance of publicity. The requirement, however, needs immediate qualification. All governments issue threats to help ensure that their subjects comply with their directives. And a credible threat certainly gives subjects a reason to do what they are being ordered to do. So, in one respect, all governments engage their subjects’ capacity to respond to reasons. Yet to respect the rational agency of those subject to their power, governments must do more than this. They must provide justifications for their actions, and the justifications provided must engage their subjects’ capacity to respond to the moral considerations that support the government’s action.
The moral basis of public justification that I have just identified — respect for rational agency — does not distinguish public justification from the moral reasons view of political justification. Proponents of the latter view can insist that governments, by offering sound moral reasons in public for the actions they undertake, respect the rational agency of their subjects. But respect for rational agency, the proponent of public justification will insist, is not realized merely by presenting persons with moral reasons to accept governmental actions. You can present a person with a reason to accept a course of action, but you will not engage his capacity to respond to reasons if he is not positioned to appreciate the force of the reason that you have given him. A distinction is helpful here. Let us say that a valid reason is a reason that applies to a person whether or not he appreciates, or even can appreciate, it. And let us say that an acceptable reason is a valid reason that a person is positioned to appreciate. Footnote 9 The ideal of public justification holds that governments must publicly justify their actions and that the justifications offered must provide their subjects with acceptable reasons for those actions. Footnote 10
The requirement to provide acceptable reasons, as opposed to merely valid reasons, explains the relational dimension of public justification. For, presumably, different subjects can be differently positioned to appreciate valid reasons; and so a successful public justification offered to one may not be a successful public justification when offered to another. But the requirement here is also puzzling. Why is a valid reason that has been presented to a subject not also an acceptable reason for him? We need to know more about what it means to say that a person is not positioned to appreciate a valid moral reason that has been presented to him in public by his government. If acceptable reasons must be indexed to the position of subjects, then we need to identify the features of subjects that fix their position.
A first response to this issue can be quickly rejected. Rawls claimed that justification “is addressed to others who disagree with us, and therefore it must always proceed from some consensus, that is, from premises that we and others publicly recognize as true.” Footnote 11 This makes justification too dependent on what people actually accept. What Rawls should have said, and what he probably meant to say, is that justification in politics should proceed from premises that we and others, if rational, can recognize as true. The hard question is how to identify the sense of possibility expressed in this demand.
The notion of rationally possible assent can be spelled out in different ways. To say that a designated political arrangement is justified to a person is to say that she would not rationally reject it if she were in a certain state or condition, even though she may not actually be in, or ever be in, that state or condition. But the hypothesized state or condition cannot be one in which the person appreciates all the valid moral reasons that apply to the justification of the political arrangement in question. For in claiming this we would, in effect, efface the distinction between public justification and the moral reasons view of political justification. Public justification is justification to people, in some respect, as they are. That is why it is indexed to them. But, at the same time, public justification is rational justification, not mere assent. That is why it must not simply take people as they are, but proceed from some idealization of them.
II. Two Types of Justifying Reasons
In the next section I will consider a range of indexing proposals that could inform the ideal of public justification. These proposals idealize the subjects of justification. A question I will be asking is whether there is a nonarbitrary point for stopping the process of idealization. But first, in this section, I will discuss an important distinction between types of justifying reasons. This distinction will help with the evaluation of the indexing proposals I will examine; and, as will emerge, the distinction is crucial for understanding the content of the requirement to respect the rational agency of persons.
The ideal of public justification holds that state officials must publicly justify political arrangements to those subject to them and that the justifications offered must provide these subjects with acceptable reasons to accept the arrangements. To know what an acceptable reason is in this context we need to understand the notion of reason that is being employed. Compare these two claims.
-
(1) State officials present in public a political justification for a designated political arrangement that invokes a consideration that, with respect to a particular subject, establishes that there is sufficient moral reason for him to support the arrangement.
-
(2) State officials present in public a political justification for a designated political arrangement that invokes a consideration that, with respect to a particular subject, establishes that there is sufficient reason for him to believe that there is sufficient moral reason for him to support the arrangement.
The consideration mentioned in (1) is a fact that explains why the political arrangement in question is morally justified. The consideration mentioned in (2) is evidence for the existence of such a fact. Henceforth, I will refer to considerations of the first kind as Type (1) reasons and considerations of the second kind as Type (2) reasons. Footnote 12
I earlier distinguished a valid reason from an acceptable reason. All acceptable reasons are valid reasons, but not all valid reasons are acceptable reasons to all subjects. We now can see two different ways in which a public justification could present a valid and acceptable reason for some subjects while failing to do so for others. The public justification might fail to do so because it appeals to Type (1) reasons that apply to some subjects, but not to others; and/or it might fail to do so because it appeals to Type (2) reasons that some subjects have but others do not. An example illustrates the distinction. Suppose some subjects promised to accept the outcome of a certain political procedure. The fact of their promise could make it true that they should support the decisions reached by the procedure. But this consideration obviously would not apply to others who had made no such promise. The promisors would have Type (1) reasons to support the decisions of the procedures that the non-promisors would not have. Now imagine that the promisors promised to accept the outcome of the procedure only on the condition that the procedure was executed fairly. And, imagine further, that some of the promisors have compelling evidence to believe that the procedure was not executed fairly (even though in reality it was executed fairly), while others have no such evidence. Given this scenario, the first group of promisors would not have the Type (2) reason to support the outcome that the second group would have. A valid reason to support a political arrangement could in this way fail to be an acceptable reason for some because they had no reason to believe in its existence.
We need to keep these distinctions in mind as we move forward. Successful public justifications are indexed to subjects; but an indexing proposal could accent either Type (1) or Type (2) reasons. Put differently, respecting the rational agency of a subject could be said to be a matter of respecting his capacity to respond to the moral reasons that apply to him or a matter of respecting his capacity to respond to the epistemic reasons that concern the moral reasons that apply to him. In due course it will become apparent that the best prospect for distinguishing public justification from the moral reasons view of political justification foregrounds Type (2) as opposed to Type (1) reasons.
III. Preliminary Proposals
This section considers four indexing proposals that purport to ground the relational dimension of public justification. Each of the proposals is inadequate. The critical discussion of them serves two purposes. First, it will help to distinguish the view I am presenting — the pure view — from some other views in the literature on public justification. Second, it will put us in a better position to see the contours of a successful indexing proposal.
A. Motivational infirmities
Recall that we are seeking to understand what it means to say that a person is positioned to appreciate the valid reasons that support a given political arrangement. Speaking generally, garden-variety irrationality of various sorts can obstruct our response to the reasons we have. Sometimes we are irrational in the sense that we do not do what we judge we have most reason to do or we do not believe what we judge we have conclusive evidence for believing. For example, a man may know that he should stop smoking, but he may have no motivation to do so because he is in the grip of an addiction. Or, to take another example, a woman may know that she should begin to look for a job, but she may be too depressed to do so. These humdrum examples show how a person can judge from his own point of view that he has decisive reason to do something, or to believe that he should do something, and yet fail to respond appropriately.
There are political analogues to the humdrum examples. Suppose a citizen knows that those who govern him have won the election in a fair contest, and he believes correctly that victory in such contests entitles the victor to govern; but he cannot bring himself to accept the legitimacy of the rulers. Seething with anger, he refuses to accept what, from his own standpoint, he recognizes he has good reason to accept. Plainly, in such a case, the rulers’ claim to rule has been justified to him.
Reflection on these examples suggests a very minimal condition on public justification. For a political arrangement to be justified to a person, it must be the case that the person would not reject the arrangement in the absence of psychological compulsions, emotional disturbances, and related motivational infirmities. No proponent of public justification of whom I am aware rejects this minimal condition.
B. Stipulated Commitments
A second indexing proposal looks to have more bite. It first stipulates that the subjects of public justification have certain moral commitments and then holds that for a political arrangement to be justified to a person it must be the case that she would have reason to accept it, given that she has these moral commitments. This proposal has proven to be a popular one. And it is easy to see why. For one thing, it makes it easier to show that certain political arrangements are publicly justified. But, more interestingly, it may seem to follow from the underlying basis of public justification itself. Following others, I have said that public justification is grounded in a notion of respect for rational agency; and I have said further that this notion implies that governments that take public justification seriously must engage their subjects’ capacities to respond to the moral reasons that support the government’s actions. But this demand presupposes that these subjects can appreciate and respond appropriately to moral reasons. This presupposition, in turn, makes sense only if subjects are understood to be committed to morality as a source of reasons. The subjects of public justification must, in short, acknowledge the authority of morality.
Not too much follows from this point, however. To explain: consider two examples of moral commitments that are often attributed to the subjects of public justification.
Equality: All persons are of equal moral standing.
Agreement: Each person is committed to living under political arrangements that all others who are subject to them can reasonably accept, given that these others too have this commitment.
Equality is sometimes taken to be a moral platitude affirmed by anyone who has an adequate grasp of the concept of morality. Footnote 13 Even if we grant this claim, it does not follow that anyone who has an adequate grasp of the concept of morality must have a correct view about the best conception of the abstract claim expressed in Equality. Equality conflicts with various elitist doctrines about the standing of persons. So it has some critical force. But it grounds only a fairly minimal stipulated commitment. The situation with Agreement is more transparent. Agreement articulates a contractualist norm. But, even if contractualism were the correct view of morality (something which I do not believe) it would not follow that anyone who recognized the authority of morality must be a contractualist. Obviously, people who are committed to morality can have false beliefs about its nature and content.
So, like the proposal concerning motivational infirmities, the stipulated commitments proposal does not take us too far in the process of idealizing the subjects of public justification. Appealing to commitments like Equality (under some nonrationally mandated interpretation) and Agreement may make it easier to achieve the ideal of public justification. Those who reject these commitments are sometimes said to fall outside the ambit of public justification. The commitments in question, however, are stipulated commitments, since, on the proposal under consideration, not all persons are rationally committed to them. (Note that if all persons are rationally committed to them, then the present indexing proposal would not be an indexing proposal at all.)
Despite their popularity, appeals to stipulated commitments betray the moral basis of public justification. That basis, to recall, is the requirement to respect the rational agency of those who are subject to political arrangements. Since those who do not have the stipulated commitments are persons with rational capacities, they too are owed respect. A publicly justified arrangement, on this indexing proposal, would not be one that respected the rational agency of all who were subject to it. Nor will it help to say that those without the stipulated commitments have been given a public justification since they would have reason to accept the arrangement on the condition that they had the stipulated commitments. For, as I have just explained, stipulated commitments are not rationally required commitments.
C. Deeply held values
You may suspect that so far I have been looking in the wrong place for a plausible indexing proposal. Stipulated commitments and motivational infirmities do not account for the diversity of outlooks to which the concern for public justification responds. People in modern societies disagree on fundamental matters about the value or meaning of life. Public justification must be indexed to these deeply held beliefs, some of which may be so much a part of the self-conception of those who hold them that they could not imagine themselves without the beliefs.
Public justification, on this third proposal, is indexed to these deeply held value commitments. This proposal is often combined with the stipulated commitments proposal. On the combined proposal, public justification should be indexed to the deeply held values of citizens who are committed to moral equality and to living together with others on terms that they too can reasonably accept. But we must reject this third proposal, on either its simple or combined version. I have already explained why the stipulated commitments proposal should be rejected, and its rejection suffices to rule out the combined proposal. Given the underlying moral basis for public justification that I have proposed, the simple version of the proposal also must be rejected. The fact that a value commitment is deeply held tells us nothing about how rational the person is in affirming it. Some deeply held value commitments are deeply unreasonable, and respect for persons as rational agents does not enjoin us to respect these commitments. To the contrary, we may show more respect for a person, understood as a rational agent, if we do not treat his deeply held values as simply constitutive of who he is, but rather treat him as a being who has the capacity to distance himself from these commitments. Public justifications, accordingly, can provide people with acceptable reasons to assent to a political arrangement, even if such people would need to abandon their deeply held values to acknowledge the force of the reasons provided. Footnote 14
D. Person-dependant reasons
A fourth proposal takes a more direct approach to grounding the relational dimension of public justification. Many Type (1) reasons are shared. If some subjects have them, then all do. Appeal to shared reasons will not help us in formulating an indexing proposal. But some Type (1) reasons are not shared. I shall call them person-dependant reasons. Footnote 15 As I have already pointed out, a Type (1) reason could be grounded in the promise or voluntarily assumed commitment of an individual. Relatedly, people have different aims, projects, and relationships, and these facts about them explain how they can have different reasons to support or oppose certain political arrangements. This possibility follows from a basic fact about value. There is a difference between recognizing something as valuable and valuing it. Footnote 16 Exploring outer space is a valuable activity. I may recognize this to be the case, but I, unlike you, do not value it. I value other activities, such as the production of opera. If the state funds space exploration, then you have a reason to support its action that I do not have. Your reason to support it is that the state action will further an activity that you value. You have a person-dependent reason to support the state action that others like me do not have. Footnote 17
Public justifications for political arrangements could invoke person-dependent reasons. This possibility explains how a public justification could succeed when presented to you, but not succeed when presented to me. But care must be taken in describing cases of this sort. Recall that public justifications must provide moral reasons to support political arrangements. Without further discussion we cannot assume that when there is a person-dependant reason for you to support a given political arrangement there is a moral reason for you to do so. For consider: in supporting your projects, the state may be unfairly favoring you over others who have an equal claim to have their projects supported. Suppose the state funds space exploration instead of opera and that its decision is unfair to the opera lovers. Then your person-dependent reason to support the state is not a moral reason to support it and it could not figure in a successful public justification for the state’s action. If, alternatively, the state funds space exploration after reaching this decision in a way that is fair to all, then all its subjects have a reason, one grounded in fairness, to support its action. You would have an additional person-dependent reason to do so; but when the state publicly justifies its action it would need to appeal to the fairness-based reason, and this is a reason that all its subjects have, if any of them do.
In sum, to be admissible in public justification, these person-dependent reasons must be accompanied by reasons that are not person-dependent. The reasons that are not person-dependent are the reasons that justify the state’s action in furthering the person-dependent goals and activities of some of its subjects, but not others. Proposals that purport to capture the relational dimension of public justification by appeal to person-dependent reasons, accordingly, look to be unpromising.
IV. Epistemic Exemplars
The preceding discussion was intended to clear the field of various proposals for giving content to the ideal of public justification — proposals that aim to capture the relational dimension of public justification and thereby distinguish it from the moral reasons view of political justification. I want to move now from criticism to construction. To identify an indexing proposal with promise, we need to foreground Type (2) reasons, since Type (1) reasons — bracketing the person-dependent reasons just discussed — do not vary across citizens. Type (2) reasons are relativized to the evidential states of those to whom they apply. They also may be relativized to the standards of reasoning that people employ to draw inferences from the evidence available to them. Both of these claims require elaboration. However, once clarified, they will give us a new indexing proposal, one that ties public justification to the rationally exemplary, yet divergent, judgments of political subjects.
In discussing motivational infirmities, we saw the need to idealize the subjects of public justification. We imagined them free of psychological compulsions, emotional disturbances, and the like. But idealization must go further. It must apply not only to mismatches between a person’s judgments and his responses to those judgments, but also to the judgments themselves. One important way in which a judgment can be defective is by being false. The idea that I want to explore is that public justification is indexed to the exemplary beliefs of the subjects of public justification. As will emerge, this idea licenses a very good measure of idealization, but it allows that exemplary judgments can be false. It must allow this if it is to capture the relational dimension of public justification and avoid collapsing into the moral reasons view of political justification.
Exemplary epistemic agents, I will argue, exhibit skill with respect to a number of tasks. Footnote 18 These tasks are discussed below. But first I need to caution against a misunderstanding. Deficits in skill can be due to the fault of the agent, but they need not be. I may be a poor golfer, even if my bad swing is not rooted in my failure to practice or my failure to seek out appropriate training. Indeed, I might have behaved as well as anyone could have in my efforts to excel at golf and yet I remain poor at golf. Conscientious epistemic agents, that is, agents who fulfill all their epistemic duties and responsibilities, can fail to be exemplary epistemic agents. For this reason, I will not speak of fault, but of defects, even though some defects are indeed due to the fault of the agent.
A. Judgments
One task an epistemic agent confronts concerns the issue of whether he should form a belief on some particular subject, given that he has the option of forbearing from doing so. This option may not always be present; the evidence (that a subject is aware of) may compel belief. But I am not concerned now with the question of the voluntariness of belief. What seems plain is that a person can believe a proposition when he should forbear from doing so. For example, the evidence available to him may have made it too risky for him to hazard a belief, but he may have done so anyway.
So we can take an actual subject and idealize her so that she only forms beliefs, as opposed to forbearing from doing so, when it is appropriate for her to do so. An exemplary epistemic agent does not believe too rashly or too timidly, given her concern with acquiring true beliefs and avoiding false beliefs. The skill or competence that an agent manifests in dealing with this task is a matter of degree. An exemplary epistemic agent excels at the task.
What I have just said about beliefs applies to judgments as well. Sometimes we should withhold judgment on a matter, given what we know or what evidence is available to us. Naturally, this general claim applies to political judgments, including political judgments that concern whether the political arrangements to which we are subject are justified or not. And this claim, in turn, raises an issue that the pure theory of public justification needs to address. To know whether a particular political arrangement has been justified to a particular subject we need to know whether (i) the subject, idealized to be an exemplary epistemic agent, would need to have a justified belief that the arrangement is justified, or (ii) whether instead it would suffice for such an agent, so idealized, to have no justified belief that the arrangement is unjustified. If an agent suspends judgment on the question of whether the arrangement is justified, and if he exercises exemplary skill in doing so, then it will follow that he has no justified belief that the arrangement is unjustified, even though he has no affirmative belief either that the arrangement is justified.
The issue is whether to insist on (i) or (ii). The pure theory holds that a political arrangement is not publicly justified to a person if that person, suitably idealized, would reject the arrangement, viewing it is an unjustified imposition on him. However, if such a person, suitably idealized, would not reject the arrangement — either because he would affirm it or he would withhold judgment on its justification — then the arrangement qualifies as publicly justified to him.
On reflection, this position seems like the right one. Suppose that there are valid Type (1) reasons that justify a political arrangement and that these reasons have been presented in public and that state officials have discharged the duty they have to present these reasons in public in a manner that facilitates their public justification. Footnote 19 With these suppositions in place, imagine now that a particular subject, S, suitably idealized, would withhold judgment on the question of whether this political arrangement is justified. By assumption, S’s Type (2) reasons are not robust enough to warrant forming a judgment on the matter. In this scenario, it is plausible that the arrangement is publicly justified to S, for S cannot object that the arrangement is an unjustified imposition on him. If this is correct, then ignorance — more precisely, ignorance of a certain kind — adds to, rather than detracts from, public justification.
B. Evidence
Exemplary epistemic agents sometimes withhold judgment, manifesting skill in doing so. Exercising this skill requires them to assess whether the total relevant evidence accessible to them, given their situation, justifies them in withholding judgment on the matter at hand. Let us look more closely at these notions of evidence and accessibility.
A citizen forms a judgment on the justifiability of a political arrangement to which he is subject. This judgment will be defective if it is not properly responsive to the evidence available to him. If we idealize this subject so that he is an exemplary epistemic agent, then he will not form judgments that are not adequately supported by his evidence. He will make sure that his judgments are properly responsive to his evidence. To determine the judgments that an epistemic agent of this kind would make we must identify the set of facts that constitute his evidence. We can call this set of facts the agent’s “epistemic field.” Footnote 20 Importantly, an agent can be unaware of facts in his epistemic field. There can be facts that he could have discovered if he had conducted a proper inquiry, but of which he is unaware.
Epistemic fields determine an agent’s epistemic reasons. Type (2) reasons are epistemic reasons, and so, unlike Type (1) reasons, they are relativized to epistemic fields. This relativity, in turn, explains the relational dimension of public justification on the pure theory. Different subjects of public justification have different epistemic fields. Idealizing these subjects to their epistemic exemplars leaves this relativity in place, thus explaining how some valid reasons to support or oppose political arrangements could be acceptable to some, but not to others. But now it might be asked, why should we not idealize the epistemic fields as well? If the pure theory helps itself to strong idealization with respect to the subjects of public justification, transforming them into epistemic exemplars, why should it not go further and idealize the evidence that is accessible to them? In one sense, it should. The epistemic field of an actual agent is determined in part by his rational competence. If he has limited powers to comprehend relevant evidence, then this evidence will lie outside his epistemic field. But the pure theory of public justification does not index public justification to the rational deficiencies of the subjects of public justification. It does not recommend a compromise with irrationality in that way. Instead, it invites us to think of epistemic exemplars that correspond to these subjects; and so we should, accordingly, think of epistemic fields that apply to subjects so idealized. But this is where the idealization stops. Epistemic exemplars can have access to different evidence, and this fact about them does not arise from any defect in their competence as epistemic agents.
Let the epistemic field of the epistemic exemplar of an actual subject of public justification be called a purified epistemic field (PEF). We can now say that for a political arrangement to be justified to a person it must be true that he would not have sufficient reason — Type (2) reason — to reject it as unjustified given his PEF. Just as different subjects of public justification have different epistemic fields, their exemplary epistemic counterparts have different PEFs. We should not overstate this divergence. People converse with one another. If some have more relevant information than others, then, at least in most cases, they can share this information with them. People also can defer to experts. When conducted well, political debate and deliberation pool evidence, thereby reducing the divergence in subject’s PEFs. Yet some divergence will almost certainly remain, Footnote 21 and that is all that is needed for the present proposal to succeed as an indexing proposal.
Before moving forward, I need to clarify a critical feature of the proposal. I have said that there can be facts in an agent’s epistemic field of which he is unaware, facts that he would discover only if he conducted a proper inquiry into the matter at hand. But what considerations guide an inquiry of this kind? On the pure theory, the only relevant considerations are epistemic or truth-related. Subjects can have practical reasons for not pursuing answers to certain questions. Likewise, subjects can have practical interests that militate against their pursuing certain lines of investigation. These practical considerations are relevant to an assessment of an agent’s practical rationality, but they are not relevant to an assessment of his epistemic reasons. One can have practical reasons for not discovering one’s epistemic reasons with respect to certain matters. Given its focus on Type (2) reasons, however, the pure theory of public justification must work with a pure account of epistemic rationality.
C. Standards
Public justification is indexed to subjects understood as exemplary epistemic agents. These agents manifest skill in forming and/or withholding judgments on questions that bear on the justification of the political arrangements to which they are subject. To exercise this skill, these agents must exercise skill in accessing the facts in their epistemic field. Yet once they have accessed the total relevant evidence provided by their epistemic field, a further task remains. They must respond skillfully to that evidence, at least if they are to qualify as exemplary epistemic agents.
We need to clarify the idea of responding skillfully to evidence. Consider the following thesis.
(Uniqueness) For any given body of evidence that bears on the justification of any specified political arrangement, there is a uniquely rational response to be drawn from the evidence concerning the justifiability, or lack of it, of the political arrangement. This response follows from the correct application of the correct set of standards of reasoning. Footnote 22
A set of standards of reasoning governs belief formation. Applied to politics, the norms establish a bridge between a person’s evidence and his Type (2) reasons. If the above thesis were correct, then a person would respond skillfully to his evidence only if he correctly applied the correct standards.
I do not wish to take a stand on whether Uniqueness is true. I want to allow, without endorsing, a departure from it. Suppose then that there are two or more sets of standards of reasoning that are equally reliable at transmitting truth from evidence to conclusions. If so, then even if two subjects were in exactly the same PEF, they might affirm conflicting beliefs about the justifiability of a political arrangement because each correctly employed different, but equally reliable, standards of reasoning. We can now supplement the indexing proposal advanced above. That proposal indexed public justification to the Type (2) reasons of subjects. Divergent Type (2) reasons, we can now say, are explained by reference to two factors: (i) the subjects have different PEFs, and/or (ii) they employ different, but equally reliable, standards of reasoning in drawing inferences from their evidence. (Of course, if Uniqueness is correct, then [ii] is unnecessary.)
To sum up: on the pure theory, public justifications for political arrangements are indexed to the Type (2) reasons that subjects have to support or oppose the political arrangements in question. Type (2) reasons are identified by reference to the ideal of an exemplary epistemic agent — an agent that exercises skill in forming judgments, accessing evidence, and applying standards of reasoning. Exemplary epistemic agents are highly idealized versions of actual political subjects. Many actual political subjects will not be aware of all the evidence available to them that pertains to the justification of the political arrangements that apply to them, and many will not apply correct standards of reasoning in forming judgments about the justification of these arrangements. Less demanding proposals are certainly available, as we saw in our consideration of preliminary indexing proposals. One could take the above proposal and decrease the degree of idealization either by requiring subjects to have less access to relevant information or by allowing them to apply incorrect standards of reasoning. Decreasing the degree of idealization, however, would amount to a compromise with rational epistemic justification. On the view I am proposing, public justification expresses the demand to justify political arrangements to subjects understood as exemplary agents. It does not ask state officials to tailor their justifications to the epistemic defects of actual subjects. A pure view of public justification cannot settle for anything less than this proposal.
V. An Interlude
The moral reasons view of political justification accents the considerations that count in favor of (and against) the justification of political arrangements. These are Type (1) reasons. By so doing, it fails to capture the relational dimension of public justification. By shifting attention to the Type (2) reasons to support or oppose political arrangements, we explain the distinction between valid and acceptable reasons; for a valid Type (1) reason to support an arrangement will not count as an acceptable reason for a person to support the arrangement if that person’s Type (2) reasons either do not make reference to it or undercut belief in it.
The distinction between Type (1) and Type (2) reasons also brings into view two dimensions of rational agency. One is the capacity to respond appropriately to the reasons we have to support or oppose political arrangements; the other is the capacity to form exemplary judgments about the reasons we have to support or oppose these arrangements. The moral reasons view of political justification holds that we respect the rational agency of persons when we respond well to the first dimension of their rational agency. The pure view of public justification holds that we respect the rational agency of persons when we respond well to the second dimension of their rational agency.
Having distinguished the pure view of public justification from the moral reasons view of political justification, we can ask how far apart the two views really are. We do not tend to think that moral reasons in general, and reasons of political morality in particular, are esoteric. It is a governing assumption of our practice of political and moral debate that valid reasons for supporting or opposing political arrangements are considerations that everyone can appreciate. That optimistic assumption directs us to think that if we believe that a valid reason exists to support an arrangement, then we should believe that that reason is in principle accessible to others. No rational agent, on the governing assumption, is epistemically cut off from the reason in question.
The governing assumption is too strong. Given his epistemic position, a person could fail to have access to a valid consideration that counts in favor of a given political arrangement, even if his performance as an epistemic agent were exemplary. Some may think that we can eliminate divergence in our exemplary judgments by a kind of retreat. Call strong modesty the view that whenever two or more epistemically exemplary agents would disagree on whether a given political arrangement is justified, then all should abandon their convictions that their respective judgments are correct. Each should modify his view to bring it into line with the others.
Neither the governing assumption nor strong modesty is satisfactory, although both express part of the truth. Too much optimism in the face of apparent defect-free divergence exaggerates the extent to which people can share their evidence. Too much modesty in the face of this disagreement gives too much weight to the views of others. Footnote 23 Still, both responses, properly understood, help us to appreciate the nature and significance of the gap between pure public justification and the moral reasons view. While strong modesty is excessive, some modesty is called for when one is confronted with apparent defect-free disagreement. One should lower one’s confidence in one’s judgments if one can identify no mistake that one’s opponents are making and one has no good noncircular explanation for why they are in error. At the same time, while too much optimism is unwarranted, there is nonetheless something right about the governing assumption. We treat others with respect when we treat them as capable of responding well to the reasons that apply to them.
VI. The Complexity of Respect
The truth about respect, as it relates to politics, is complicated, however. There are, I now want to suggest, two demands of respect for rational agency that correspond to the two dimensions of rational agency that I have distinguished. The two demands are implicit in the slogan that “justice must not only be done, but be seen to be done.” Footnote 24 The first part of the slogan expresses one demand of respect for persons as rational agents, a demand that directs us to treat them in accord with sound principles of political morality. In my terms, this means that we must subject them to political arrangements that are justified by the moral reasons — the Type (1) moral reasons — that bear on their justification. The second part of the slogan expresses a different demand of respect. This demand is to treat them in ways that they can see to be justified. In my terms, this means that we must treat them in ways that they could see as justified given the evidence in their epistemic field and given that they have functioned as exemplary epistemic agents.
The two demands of respect potentially conflict. They may conflict because one demand directs us to the moral reasons view of political justification and the other directs us to the pure public justification view. And, as pointed out above, political arrangements can be justified by the moral reasons view, but publicly unjustified and vice versa. Possible conflicts between the demands of respect must be taken seriously if we think, as I have argued, that there is indeed a gap between the moral reasons view and the pure theory of public justification. But the possibilities in question cannot be described straightforwardly. For notice that, if the pure theory of public justification is sound, then there is a Type (1) moral reason to make sure that persons are not subjected to political arrangements that they cannot see to be justified. So let us call that particular Type (1) moral reason the Type (1) reason of public justification.
Possibly, there are political arrangements that are justified by the full set of Type (1) moral reasons, absent the Type (1) reason of public justification, but that are not justifiable to all who are subject to them in ways that they could come to see as justified given the evidence that they have available to them and given that they correctly employ reliable standards of reasoning. In response to this possibility, one might say that the Type (1) reason of public justification is a decisive reason. Once added to the mix, it dominates the other Type (1) considerations. This response eliminates the conflict between the two demands of respect. However, there is, I think, no good case for assigning the Type (1) reason of public justification this kind of decisive weight. And, if it does not have decisive weight, then the possible conflict between the two demands of respect reemerges.
VII. Application
Several remaining issues need to be addressed. The first of which has just been broached. If the moral reasons view and the pure public justification view diverge on what they justify, which view should be privileged in practice? We could refuse to answer the question. There are two kinds of justification, and a political arrangement could be justified on one, but unjustified on the other. And that, we could say, just is the end of the story. But this response would not tell us whether the arrangement in question is, all things considered, justified and this is something we may like to know.
The question can be sharpened. Public justification is relational. The same arrangement could be publicly justified with respect to A, but not publicly justified with respect to B. Proponents of public justification often claim that the ideal demands that political arrangements be publicly justified to all who are subject to them. We need not follow this practice. We can say that the extent to which a political arrangement is publicly justified is a matter of degree. On the pure theory, it is extremely unlikely that any political arrangement for a modern society could be publicly justified to all who are subject to it, given that the pure theory does not help itself to stipulated commitments. Following Rawls, most proponents of public justification have claimed that political arrangements are not legitimate if they could not be publicly justified to some who are subject to them. Such theorists tether public justification to contractualist political legitimacy. The pure theory cuts this tie. Pure public justification is an aspirational ideal, one that can be satisfied to a greater or lesser extent in modern societies, and one that should inform, but not dominate, the moral justification of political arrangements in these societies. Footnote 25
Having abandoned the quest for universal public justification, the pure theorist must settle for something more modest. We can say that a political arrangement is fully justified if it is adequately supported by Type (1) reasons and if it can be publicly justified to a wide range of those who are subject to it. With respect to those who are subject to the political arrangement in question, but to whom it is not publicly justified, we can say that the political arrangement is justifiably imposed on them, even if it is not justified to them. How, more exactly, should we characterize the normative position of those who have sufficient Type (2) reasons to reject a political arrangement that is adequately justified by Type (1) reasons? Our earlier discussion suggests that these subjects are treated with respect in one sense, but disrespect in another. They are treated respectfully insofar as they are treated as reasons-responsive agents who are subject to the applicable Type (1) reasons. They are treated with disrespect insofar as they are subjected to an arrangement that they must reject as reasons-responsive agents who have the applicable Type (2) reasons.
Confronted with this situation, two options are open to us. We may insist that Type (1) respect trumps Type (2) respect. If we opt for this first option, then pure public justification is still morally important. It adds to the full justification of political arrangements, even if it cannot justify any departure from what would be justified by the full set of relevant Type (1) reasons. The second option is to hold that the two senses of respect must be balanced against each other. On this option, we have two aims. We aim to justify political arrangements by appeal to Type (1) reasons and we aim to justify them by appeal to Type (2) reasons. And, on this second option, we should settle for achieving one aim a little less well if doing so would enable us to achieve the other aim much more substantially.
The pure theory of public justification does not tell us which of these options we must go for. It presents an account of the nature of public justification, and it tells us that this ideal is aspirational and not a constraint on political legitimacy. But it does not tell us how much weight to assign to the ideal.
This is not unexpected. It is one thing to give a theoretical account of a political ideal and another to say how exactly it should be pursued in practice. With this thought, we find a final issue that needs to be addressed. We cannot know a priori the deficiencies or shortcomings that prevent people from forming exemplary beliefs about the moral reasons that apply to the justification of the political arrangements to which they are subject. Further, whether or not particular people have formed exemplary beliefs regarding the justification of these arrangements is not open to easy inspection. How then, it is fair to ask, could state officials aim at pure public justification?
Once again, our answer is modest. State officials should not concern themselves directly with the epistemic states of those over whom they govern. They should pursue pure public justification indirectly by removing unnecessary obstacles to it. In justifying political arrangements, they should avoid unnecessary appeal to doctrines or claims that are widely disputed in the society. They should attempt to justify these arrangements by appeal to premises that are broadly, even if not universally, acceptable to those who are subject to them. In pluralistic societies, this means often abjuring appeal to deep foundations and to contested doctrines about the meaning and value of life. These views, we assume here, could be affirmed by many who are exemplary epistemic agents. To this extent, our view resembles the Rawlsian view. Footnote 26 But the reference to unnecessary appeal is crucial here. If state officials correctly believe that a political arrangement is justified on the moral reasons view, but that it cannot be adequately justified by limiting appeal to doctrines or claims that are broadly acceptable, then the appeal to doctrines that are widely disputed is not unnecessary. The appeal to these doctrines would be necessary for the adequate justification of the political arrangement in question.
This modest answer may look too modest. It holds, in effect, that pure public justification should be pursued only if it does not prevent the full and adequate justification of Type (1) justified political arrangements. It only supplements, and never supplants, the moral reasons view. But this answer seems to assume that the ideal of pure public justification is never weighty enough to justify departures from Type (1) justified political arrangements. And that strong assumption is not one that I have defended. For all that has been said here, a proponent of pure public justification could insist that the ideal vindicates political arrangements that are suboptimal from the standpoint of the moral reasons view. This is all true enough. But we are now considering how the ideal of pure public justification should be pursued in practice. In practice, state officials already have plenty of reasons to pursue broad-based political justification, reasons that have nothing to do with pure public justification. These are the reasons of prudent strategy and political stability mentioned at the outset of our investigation. Even if the pursuit of pure public justification in principle warrants departures from Type (1) justified political arrangements, in practice these departures likely will already have been vindicated by these other considerations.
Appendix
This appendix briefly discusses a complication that may apply to the distinction between the types of reasons (Type 1 and Type 2) that I have appealed to in this essay. Presumably, we can have a reason to do something, even when it is not epistemically accessible to us in our situation. But reflection on certain cases challenges this idea. Suppose that a lottery has been conducted by a computer. A number has been randomly selected by the computer, but no one knows what it is. If one now were to write that number on a particular lottery ticket and submit the ticket to the agency that conducts the lottery, then one would win the prize. Does one now have a reason to do so? Likewise, suppose a chemist can mix certain chemicals to produce a cure for a disease that has no known cure. But he has no idea, nor does anyone else, that doing so would produce this result. In these cases, facts about the inaccessibility of the putative practical reason may be thought to eliminate it.
The verdicts in these cases could be explained by introducing an agent-relative epistemic filter on practical reasons. Footnote 27 For a consideration to qualify as a reason for action it must pass through this filter. The epistemic filter would show that both Type (1) and Type (2) reasons are sensitive to epistemic factors. It might be claimed, accordingly, that the introduction of such a filter would effectively efface the distinction between Type (1) and Type (2) reasons.
This claim would be mistaken for two reasons, however. First, Type (2) reasons can rest on false beliefs about the existence of Type (1) reasons. Thus, even if it were true that all Type (1) reasons are also Type (2) reasons, it would not follow that all Type (2) reasons are Type (1) reasons. Second, the epistemic filter, while it is not well understood or subject to precise specification, is plausibly more porous than a filter that would eliminate any consideration that an exemplary epistemic agent would not be justified in accepting.
The nature of the agent-relative epistemic filter on practical reasons is, I have said, not well understood. We have no perspicuous account of it. For that reason, some will reject it. If it is rejected, then the distinction between Type (1) and Type (2) reasons is not complicated in the way I have been discussing. However, if we accept the filter, then it does not efface the distinction I have drawn. It merely complicates it.
