first-person dictatorship and the free-rider forms, since in each case a proper name, or pronoun, or a rigged definite description is needed, either to single out the dictator or to characterize the whatever, in his judgment, is most likely to further his own aims. The principle here can clearly be expressed in a perfectly general way. It is the ordering condition which renders general egoism inadmissible, for if everyone is authorized to advance his aims as he pleases, or if everyone ought to advance ha own interests, competing claims are not ranked at all and the outcome determined by force and cunning The several kinds of egoism, then, do not appear on the list presented to the parties. They are el iminated by the formal constraints. Of course, this is not a surprising conclusion, since it is obvious that by choosing one of the other conceptions the persons in the original position can do much better for themselves. Once they ask which principles all should agree to, no form of egoism is a serious candidate for consideration in any case. This only confirms what we knew already, namely, that al though egoism is logically consistent and in this sense not irrational, it is incompatible with what we intuitively regard as the moral point of view. The significance of egoism philosophically is not as an al temative conception of right but as a challenge to any such conception. In justice as fairness this is reflected in the fact that we can interpret general egos as the no-agreement Point. It is what the parties would be stuck with if they were unable to reach an understanding 24. THE VEIL OF IGNORANCE The idea of the original position is to set up a fair procedure so that any principles agreed to will be just. The aim is to use the notion of pure procedural justice as a basis of theory. Somehow we must nullify effects of specific contingencies which put men at odds and tempt them to exploit social and natural circumstances to their own advantage. now in order to do this i assume that the parties are situated behind a veil of ignorance. They do not know how the various alternatives will affect their own particular case and they are obliged to evaluate principles solely on the basis of general considerations. ll It is assumed, then, that the parties do not know certain kinds of particular facts. First of all, no one knows his place in society, his class position or social status; nor does he know his fortune in the distribution of natural assets and abilities, his inte lligence and strength, and the like. nor gain, does anyone know his conception of the good, the particulars of his rational plan of life, or even the special features of his psychology such as his aversion to risk or liability to Optimism or pessimism. More than this, I assume that the parties do not know the particular circumstances of their own society. That is, they do not know its economic or pol itical situation, or the level of civilization and culture it has been able to achieve. The persons in the original position have no information as to which generation they belong. These broader restrictions on knowledge are appropriate in part because questions of social justice arise between generations as well as within them, for example, he question of the appropriate rate of capital saving and of the conservation of natural resources and the environment of nature There is also, theoretically anyway, the question of a reasonable genetic policy. In these cases too, in order to carry through the idea of the original position, the parties must not know the contingencies that set them in opposition. They must hoose principles the consequences of which they are prepared to live with whatever generation they turn out to belong to As far as possible, then, the only particular facts which the parties know is that their society is
first-person dictatorship and the free-rider forms, since in each case a proper name, or pronoun, or a rigged definite description is needed, either to single out the dictator or to characterize the freerider. Generality does not, however, exclude general egoism, for each Person is allowed to do whatever, in his judgment, is most likely to further his own aims. The princip1e here can clearly be expressed in a perfectly general way. It is the ordering condition which renders general egoism inadmissible, for if everyone is authorized to advance his aims as he pleases, or if everyone ought to advance ha own interests, competing claims are not ranked at a1l and the outcome is determined by force and cunning. The several kinds of egoism, then, do not appear on the list presented to the parties. They are e1iminated by the formal constraints. Of course, this is not a surprising conclusion, since it is obvious that by choosing one of the other conceptions the persons in the original position can do much better for themselves. Once they ask which princip1es all should agree to, no form of egoism is a serious candidate for consideration in any case. This only confirms what we knew already, namely, that a1though egoism is logically consistent and in this sense not irrational, it is incompatible with what we intuitively regard as the moral point of view. The significance of egoism philosophically is not as an a1temative conception of right but as a chal1enge to any such conception. In justice as fairness this is reflected in the fact that we can interpret genera1 egoism as the no-agreement Point. It is what the parties would be stuck with if they were unable to reach an understanding. 24. THE VEIL OF IGNORANCE The idea of the original position is to set up a fair procedure so that any principles agreed to will be just. The aim is to use the notion of pure procedural justice as a basis of theory. Somehow we must nullify effects of specific contingencies which put men at odds and tempt them to exploit social and natural circumstances to their own advantage. Now in order to do this I assume that the parties are situated behind a veil of ignorance. They do not know how the various alternatives wi1l affect their own particu1ar case and they are obliged to evaluate principles solely on the basis of general considerations.11 It is assumed, then, that the parties do not know certain kinds of particular facts. First of all, no one knows his place in society , his class position or social status; nor does he know his fortune in the distribution of natural assets and abi1ities, his inte1ligence and strength, and the like. Nor, again, does anyone know his conception of the good, the particulars of his rational plan of life, or even the special features of his psychology such as his aversion to risk or liability to Optimism or pessimism. More than this, I assume that the parties do not know the particu1ar circumstances of their own society. That is, they do not know its economic or political situation, or the level of civilization and culture it has been able to achieve. The persons in the origina1 position have no information as to which generation they belong. These broader restrictions on know1edge are appropriate in part because questions of social justice arise between generations as well as within them, for example, he question of the appropriate rate of capital saving and of the conservation of natural resources and the environment of nature There is also, theoretically anyway, the question of a reasonable genetic policy. In these cases too, in order to carry through the idea of the origina1 position, the parties must not know the contingencies that set them in opposition. They must choose principles the consequences of which they are prepared to live with whatever generation they turn out to belong to. As far as possible, then, the only particular facts which the parties know is that their society is
subject to the circumstances of justice and whatever this implies It is taken for granted, however that they know the general facts about human society. They understand political affairs and the principles of economic theory they know the basis of social organization and the laws of human psychology. Indeed, the parties are presumed to know whatever general facts affect the choice of the principles of justice. there are I1. The veil of ignorance is so natural a condition that something like it must have occurred to many. The closest explicit statement of it known to me is found in J. C. Harsanyi, " Cardinal Utility in Welfare Economics and in the Theory of Risk-Taking, " Journal of Political Economy, vol, 61(1953). Harsanyi uses it to develop a utilitarian theory, as I discuss below in $$27--28 no limitations on general information, that is, on general laws and theories, since conceptions of justice must be adjusted to the characteristics of the systems of social cooperation which they are to regulate, and there is no reason to rule out these facts, It is, for example, a consideration against a conception of justice that, in view of the laws of moral psychology, men would not acquire a desire to act upon it even when the institutions of their society satisfied it. For in this case there would be difficulty in securing the stability of social cooperation. It is an important feature of a conception of justice that it should generate its own support. That is, its principles should be such that when they are embodied in the basic structure of society men tend to acquire the corresponding sense of justice. Given the principles of moral learning, men develop a desire to act in accordance with its principles. In this case a conception of justice is stable. This kind of general information is admissible in the original position. The notion of the veil of ignorance raises several difficulties. Some may object that the exclusion of nearly all particular information makes it difficult to grasp what is meant by the original position. Thus it may be helpful to observe that one or more persons can at any time enter this position, or perhaps, better, simulate the deliberations of this hypothetical situation, simply by reasoning in accordance with the appropriate restrictions. In arguing for a conception of justice we must be sure that it is among the permitted alternatives and satisfies the stipulated formal constraints. No considerations can be advanced in its favor unless they would be rational ones for us to urge were we to lack the kind of knowledge that is excluded The evaluation of principles must proceed in terms of the general consequences of their public ecognition and universal application, it being assumed that they will be complied with by everyone. To say that a certain conception of justice would be chosen in the original position is equivalent to saying that rational deliberation satisfying certain conditions and restrictions would reach a certain conclusion. If necessary, the argument to this result could be set out more formally I shall, however, speak throughout in terms of the notion of the original position It is more economical and suggestive, and brings out certain essential features that otherwise one might easily overlook marks show that the original position is not to be thought of as a general assembly whicl at one moment everyone who will live at some time, or, much less, as an assembly of everyone who could live at some time. It is not a gathering of all actual or possible persons. To conceive of the original position in either of these ways is to stretch fantasy too far; the conception would cease to be a natural guide to intuition. In any case, it is important that the original position be interpreted so that one can at any time adopt its perspective. It must make no difference when one takes up this viewpoint, or who does so the restrictions must be such that the same principles
subject to the circumstances of justice and whatever this imp1ies. It is taken for granted, however that they know the genera1 facts about human society. They understand political affairs and the princip1es of economic theory ;;they know the basis of social organization and the laws of human psychology. Indeed, the parties are presumed to know whatever general facts affect the choice of the principles of justice.. there are l1. The veil of ignorance is so natural a condition that something like it must have occurred to many. The closest explicit statement of it known to me is found in J. C. Harsanyi, "CardinaI Utility in Welfare Economics and in the Theory of Risk-Taking," Journal of Political Economy, vol, 61 (I953). Harsanyi uses it to develop a utilitarian theory, as I discuss below in $$27--28. no limitations on general information, that is, on general laws and theories, since conceptions of justice must be adjusted to the characteristics of the systems of social cooperation which they are to regulate, and there is no reason to rule out these facts, It is, for example, a consideration against a conception of justice that, in view of the laws of moral psychology, men would not acquire a desire to act upon it even when the institutions of their society satisfied it. For in this case there would be difficulty in securing the stability of social cooperation. It is an important feature of a conception of justice that it should generate its own support. That is, its principles should be such that when they are embodied in the basic structure of society men tend to acquire the corresponding sense of justice. Given the principles of moral learning, men develop a desire to act in accordance with its principles. In this case a conception of justice is stable. This kind of general information is admissible in the original position. The notion of the veil of ignorance raises several difficulties. Some may object that the exclusion of nearly all particular information makes it difficult to grasp what is meant by the original position. Thus it may be helpful to observe that one or more persons can at any time enter this position, or perhaps, better, simulate the deliberations of this hypothetical situation, simply by reasoning in accordance with the appropriate restrictions. In arguing for a conception of justice we must be sure that it is among the permitted alternatives and satisfies the stipulated formal constraints. No considerations can be advanced in its favor unless they would be rational ones for us to urge were we to lack the kind of knowledge that is excluded. The evaluation of principles must proceed in terms of the general consequences of their public recognition and universal application, it being assumed that they will be complied with by everyone. To say that a certain conception of justice would be chosen in the original position is equivalent to saying that rational deliberation satisfying certain conditions and restrictions would reach a certain conclusion. If necessary, the argument to this result could be set out more formally. I shall, however, speak throughout in terms of the notion of the original position. It is more economical and suggestive, and brings out certain essential features that otherwise one might easily overlook. 138 These remarks show that the origina1 position is not to be thought of as a general assembly which includes at one moment everyone who will live at some time; or, much less, as an assembly of everyone who could live at some time. It is not a gathering of all actual or possible persons. To conceive of the original position in either of these ways is to stretch fantasy too far; the conception would cease to be a natural guide to intuition. In any case, it is important that the original position be interpreted so that one can at any time adopt its perspective. It must make no difference when one takes up this viewpoint, or who does so: the restrictions must be such that the same principles
are always chosen. The veil of ignorance is a key condition in meeting this requirement. It insures not only that the information available is relevant, but that it is at all times the same It may be protested that the condition of the veil of ignorance is irrational. Surely, some may object principles should be chosen in the light of all the knowledge available. There are various replies to this contention. Here I shall sketch those which emphasize the simplifications that need to be made if one is to have any theory at all. (Those based on the Kantian interpretation of the original position are given later, $40. ) To begin with, it is clear that since the differences among the parties are unknown to them, and everyone is equally rational and similarly situated, each is convinced by the same arguments. Therefore, we can view the choice in the original position from the standpoint of one person selected at random. If anyone after due reflection prefers a conception of justice to another, then they all do, and a unanimous agreement can be reached We can, to make the circumstances more vivid, imagine that the parties are required to communicate with each other through a referee as intermediary, and that he is to announce which alternatives have been suggested and the reasons offered in their support. He forbids the attempt to form coalitions, and superfluous, assuming that the deliberations of the parties must be simile uch a referee is actually he informs the parties when they have come to an understanding. But Thus there follows the very important consequence that the parties have no basis for bargaining the usual sense. No one knows his situation in society nor his natural assets, and therefore no one is in a position to tailor principles to his advantage. We might imagine that one of the contractees threatens to hold out unless the others agree to principles favorable to him. But how does he know which principles are especially in his interests? The same holds for the formation of coalitions: if a group were to decide to band together to the disadvantage of the others, they would not know how to favor themselves in the choice of principles. Even if they could get everyone to agree to their proposal, they would have no assurance that it was to their advantage, since they cannot identify themselves either by name or description. The one case where this conclusion fails is that of saving. since the persons in the original position know that they are contemporaries ( taking the present time of entry interpretation ) they can favor their generation by refusing to make any sacrifices at all for their successors; they simply acknowledge the principle that no one has a duty to save for posterity Previous generations have saved or they have not; there is nothing the parties can now do to affect that. So in this instance the veil of ignorance fails to secure the desired result. therefore I resolve the question of justice between generations in a different way by altering the motivation assumption. But with this adjustment no one is able to formulate principles especially designed to dvance his own cause. Whatever his temporal position, each is forced to choose for everyone. 12 The restrictions on particular information in the original position are, then, of fundamental ce. Without them we would not be able to work out any definite theory of justice at all We would have to be content with a vague formula stating that justice is what would be agreed to without being able to say much, if anything, about the substance of the agreement itself. The formal constraints of the concept of right, those applying to principles directly, are not sufficient for our purpose. The veil of ignorance makes possible a unanimous choice of a particular conception of justice. Without these limitations on knowledge the bargaining problem of the original position would be hopelessly complicated. Even if theoretical ly a solution were to exist, we would not, at present anyway, be able to determine it
are always chosen. The veil of ignorance is a key condition in meeting this requirement. It insures not only that the information available is relevant, but that it is at all times the same It may be protested that the condition of the veil of ignorance is irrational. Surely, some may object, principles should be chosen in the light of all the knowledge available. There are various replies to this contention. Here I shall sketch those which emphasize the simplifications that need to be made if one is to have any theory at all. (Those based on the Kantian interpretation of the original position are given later, $40.) To begin with, it is clear that since the differences among the parties are unknown to them, and everyone is equally rational and similarly situated, each is convinced by the same arguments. Therefore, we can view the choice in the original position from the standpoint of one person selected at random. If anyone after due reflection prefers a conception of justice to another, then they all do, and a unanimous agreement can be reached. We can, to make the circumstances more vivid, imagine that the parties are required to communicate with each other through a referee as intermediary, and that he is to announce which alternatives have been suggested and the reasons offered in their support. He forbids the attempt to form coalitions, and he informs the parties when they have come to an understanding. But such a referee is actually superfluous, assuming that the deliberations of the parties must be similar. Thus there follows the very important consequence that the parties have no basis for bargaining in the usual sense. No one knows his situation in society nor his natural assets, and therefore no one is in a position to tailor principles to his advantage. We I39 might imagine that one of the contractees threatens to hold out unless the others agree to principles favorable to him. But how does he know which principles are especially in his interests? The same holds for the formation of coalitions: if a group were to decide to band together to the disadvantage of the others, they would not know how to favor themselves in the choice of principles. Even if they could get everyone to agree to their proposal, they would have no assurance that it was to their advantage, since they cannot identify themselves either by name or description. The one case where this conclusion fails is that of saving. since the persons in the original position know that they are contemporaries (taking the present time of entry interpretation ), they can favor their generation by refusing to make any sacrifices at all for their successors; they simply acknowledge the principle that no one has a duty to save for posterity. Previous generations have saved or they have not; there is nothing the parties can now do to affect that. So in this instance the veil of ignorance fails to secure the desired result. Therefore l resolve the question of justice between generations in a different way by altering the motivation assumption. But with this adjustment no one is able to formulate principles especially designed to advance his own cause. Whatever his temporal position, each is forced to choose for everyone.12 The restrictions on particular information in the original position are, then, of fundamental importance. Without them we would not be able to work out any definite theory of justice at all. We would have to be content with a vague formula stating that justice is what would be agreed to without being able to say much, if anything, about the substance of the agreement itself. The formal constraints of the concept of right, those applying to principles directly, are not sufficient for our purpose. The veil of ignorance makes possible a unanimous choice of a particular conception of justice. Without these limitations on knowledge the bargaining problem of the original position would be hopelessly complicated. Even if theoretically a solution were to exist, we would not, at present anyway, be able to determine it
The notion of the veil of ignorance is implicit, I think, in Kant's 12. Rouaseau, The Social Contract, bk Il, ch IV, par. 5 ethics($40). Nevertheless the problem of defining the knowledge of the parties and of characterizing the alternatives open to them has often been passed over, even by contract theories Sometimes the situation definitive of moral deliberation is presented in such an indeterminate way that one cannot ascertain how it will turn out. Thus Perry s doctrine is essentially contractarian he holds that social and personal integration must proceed by entirely different principles, the latter by rational prudence, the former by the concurrence of persons of good will. He would appear to reject utilitarianism on much the same grounds suggested earlier: namely, that it improperly extends the principle of choice for one person to choices facing society. The right course of action is characterized as that which best advances social aims as these would be formulated by reflective agreement given that the parties have full knowledge of the circumstances and are moved by a benevolent concern for one anothers interests. No effort is made, however to specify in any precise way the possible outcomes of this sort of agreement Indeed. without a far more elaborate account. no conclusions can be drawn. 13 I do not wish here to criticize others; rather, I want to explain the necessity for what may seem at times like so many irrelevant details Now the reasons for the veil of ignorance go beyond mere simplicity. We want to define the original position so that we get the desired solution. If a knowledge of particulars is allowed, then the outcome is biased by arbitrary contingencies. As already observed, to each according to his threat advantage is no a principle of justice. If the original position is to yield agreements that are just, the parties must be fairly situated and treated equally as moral persons. The arbitrariness of the world must be corrected for by adjusting the circumstances of the initial contractual situation Moreover, if in choosing principles we required unanimity even when there is full infomation, only a few rather obvious cases could be decided. A conception of justice based on unanimity in these circumstances would indeed be weak and trivial. but once knowledge is excluded. the requirement of unanimity is not out of Place and the fact that 13. See R B. Perry, The General Theory of Value(New York, Longmans Green and company, 1926), PP. 674--682 it can be satisfied is of great importance. It enables us to say of the preferred conception of justice that it represents a genuine reconciliation of interests A final comment. For the most part I shall suppose that the parties possess all general information. No general facts are closed to them. I do this mainly to avoid complications. Nevertheless a conception of justice is to be the public basis of the terms of social cooperation. Since common understanding necessitates certain bounds on the complexity of principles, there may likewise be limits on the use of theoretical knowledge in the original position. Now clearly it would be very difficult to classify and to grade for complexity the various sorts of general facts. I shall make no attempt to do this. We do however recognize an intricate theoretical construction when we meet one. Thus it seems reasonable to say that other things equal one conception of justice is to be preferred to another when it is founded upon markedly simpler general facts, and its choice does not depend upon elaborate calculations in the light of a vast array of theoretically defined
The notion of the veil of ignorance is implicit, I think, in Kant's l2. Rouaseau, The SociaI Contract, bk. II, ch. IV, par. 5. 140 ethics ($40). Nevertheless the problem of defining the knowledge of the parties and of characterizing the alternatives open to them has often been passed over, even by contract theories. Sometimes the situation definitive of moral deliberation is presented in such an indeterminate way that one cannot ascertain how it will turn out。 Thus Perry's doctrine is essentially contractarian: he holds that social and personal integration must proceed by entirely different principles, the latter by rational prudence, the former by the concurrence of persons of good will. He would appear to reject utilitarianism on much the same grounds suggested earlier: namely, that it improperly extends the principle of choice for one person to choices facing society. The right course of action is characterized as that which best advances social aims as these would be formulated by reflective agreement given that the parties have full knowledge of the circumstances and are moved by a benevolent concern for one another's interests. No effort is made, however to specify in any precise way the possible outcomes of this sort of agreement. Indeed, without a far more elaborate account, no conclusions can be drawn. 13 I do not wish here to criticize others; rather, I want to explain the necessity for what may seem at times like so many irrelevant details. Now the reasons for the veil of ignorance go beyond mere simplicity. We want to define the original position so that we get the desired solution. If a knowledge of particulars is allowed, then the outcome is biased by arbitrary contingencies. As already observed, to each according to his threat advantage is no a principle of justice. If the original position is to yield agreements that are just, the parties must be fairly situated and treated equally as moral persons. The arbitrariness of the world must be corrected for by adjusting the circumstances of the initial contractual situation. Moreover, if in choosing principles we required unanimity even when there is full information, only a few rather obvious cases could be decided. A conception of justice based on unanimity in these circumstances would indeed be weak and trivial. But once knowledge is excluded. the requirement of unanimity is not out of Place and the fact that l3. See R. B. Perry, The General Theory of Value (New York, Longmans, Green and Company, l926), PP. 674--682. 141 it can be satisfied is of great importance. It enables us to say of the preferred conception of justice that it represents a genuine reconciliation of interests. A final comment. For the most part I shall suppose that the parties possess all general information. No general facts are closed to them. I do this mainly to avoid complications. Nevertheless a conception of justice is to be the public basis of the terms of social cooperation. Since common understanding necessitates certain bounds on the complexity of principles, there may likewise be limits on the use of theoretical knowledge in the original position. Now clearly it would be very difficult to classify and to grade for complexity the various sorts of general facts. I shall make no attempt to do this. We do however recognize an intricate theoretical construction when we meet one. Thus it seems reasonable to say that other things equal one conception of justice is to be preferred to another when it is founded upon markedly simpler general facts, and its choice does not depend upon elaborate calculations in the light of a vast array of theoretically defined
possibilities. It is desirable that the grounds for a public conception of justice should be evident to everyone when circumstances permit This consideration favors, I believe, the two principles of justice over the criterion of utility 25. THE RATIONALITY OF THE PARTIES I have assumed throughout that the persons in the original position are rational. In choosing between principles each tries as best he can to advance his interests. But I have also assumed that the parties do not know their conception of the good. This means that while they know that they have some rational Plan of life, they do not know the details of this plan, the particular ends and interests which it is calculated to promote. How, then, can they decide which conceptions of justice are most to their advantage? Or must we suppose that they are reduced to mere guessing? To meet this difficulty, I postulate that they accept the account of the good touched upon in the preceding chapter: they assume that they would prefer more primary social goods rather than less Of course, it may turn out once the veil of ignorance is removed, that some of them for religious or other reasons may not, in fact, want more of these goods
possibilities. It is desirable that the grounds for a public conception of justice should be evident to everyone when circumstances permit. This consideration favors, I believe, the two principles of justice over the criterion of utility. 25. THE RATIONALITY OF THE PARTIES I have assumed throughout that the persons in the original position are rational. In choosing between principles each tries as best he can to advance his interests. But I have also assumed that the parties do not know their conception of the good. This means that while they know that they have some rational Plan of life, they do not know the details of this plan, the particular ends and interests which it is calculated to promote. How, then, can they decide which conceptions of justice are most to their advantage? Or must we suppose that they are reduced to mere guessing? To meet this difficulty, I postulate that they accept the account of the good touched upon in the preceding chapter: they assume that they would prefer more primary social goods rather than less. Of course, it may turn out, once the veil of ignorance is removed, that some of them for religious or other reasons may not, in fact, want more of these goods