From now on I intend to put aside history and exegesis of texts to take up as philosophical questions some matters which arise from Bayle's argument for toleration. In fact I believe that the main conclusions I argue for in the remaining essays are substantially Bayle's, but I am not concerned to show that they are, and have not adopted them out of any loyalty to him. This third essay is an analysis of the reciprocity argument as a type. I have already discussed Bayle's version, but other versions are possible, and it seems worthwhile to analyse their common structure and consider their limitations. The fourth essay is a discussion of the ethics of belief and inquiry. This topic was touched on in connection with Fr Terrill's views on invincible ignorance (see above, Essay I, sect. 3), and again in connection with Bayle's views on culpable error, prepossession, opinionatedness and temerity (see above, Essay II, sect. 4.2). In the Philosophical Commentary see Supplement, ch. 17, "What judgement should be made of those who will not enter into disputes". But since Bayle's time a good deal has been written on the subject, and a discussion independent of his seems worthwhile. I will therefore put the texts aside and enter upon a consideration of some of the questions they have raised.
Since Bayle's time the idea of religious toleration has been developed and extended into an idea of an open or liberal society, and for the various elements of that idea many arguments have been put forward. Often these arguments do not seem to match the precise conclusions they are meant to establish. It is not enough, for example, to show that a general policy of not punishing people for what they believe will, on balance, generally have good effects. It is necessary to show that not punishing people for what they believe should be regarded as a matter of moral principle, to be followed even in the odd cases when not punishing might on balance have bad effects.Note 1 Few of the common arguments are even attempts to show that toleration should be regarded as a matter of moral principle.
Is that how toleration should be regarded? Consider the attitude of those who practised or defended what we now regard as religious persecution against the Protestants in seventeenth-century France. Most of them, Arnauld for example, were humane men who believed that conversion ought to result from persuasion, and that mildness is generally more persuasive; still, they were prepared to use coercion, together with persuasion, when they thought that that would be an effective combination; this had also been St Augustine's attitude.Note 2 This involves occasional toleration, even toleration as a (temporarily) settled policy (for example, during the time when the Edict of Nantes was substantially observed); error is tolerated while mild methods are tried, but coercive methods are kept in reserve. This is not the toleration that Bayle, Locke, Taylor and their successors contended for. What they wanted was a principled abstension from certain methods even when they might work. We may think that these methods never work, or at least are never good on balance. Then toleration means a commitment to principles ruling out such methods even in cases (if there were any) in which they might work, or otherwise be good on balance.
Principles are of different kinds. Some purport to be absolute, exceptionless and conclusive. Such principles, if there are any, are applied by simple deductive argument: "All acts of such-and-such sort are wrong and not to be done; this act I am contemplating would be of that sort; therefore it is not to be done." Others are prima facie principles.Note 3 Such a principle constitutes a prima facie case for or against some action, but it may in the end be outweighed or overridden by some reason on the other side. Prima facie principles are applied by "deliberation", that is, by "weighing" in the mind reasons on both sides of the question until the balance seems to come down on one side, until we come to feel convinced that the thing is or is not to be done. Not only moral questions but practical questions of all sorts, for example choice of a house or job, are decided by weighing up pros and cons. There may be "principles of deliberation" which specify what sorts of considerations may or may not be weighed on the other side. The prima facie principle puts a more or less weighty consideration onto one side of the balance, the "deliberative principles" (if any apply) restrict what can be put onto the other side, and the decision is that "on balance" this or that is the right or best thing to do.Note 4
If something is "a matter of principle" (and this phrase may represent some narrowing of the notion of a moral principle) then there may be cases in which what it requires is to be done even though, if it had not been a matter of principle, something else would on balance have been better. Consider the principle that promises are to be kept. Suppose I promise you something you want me to promise, in the hope that at some future date you will do something for me. If when the time comes to perform what I promised you have lost the power to be useful to me, then the hope that induced me to make the promise is no reason for performing it. But the fact that I promised is itself a reason. Suppose that in general fidelity to promises is necessary to trust among human beings, and suppose this is a strong enough reason for making it a matter of principle that promises are to be kept. In some cases it may make no difference to trust if I break a promise (suppose the person to whom I made it is dead, and no one else knows of the promise); still, the fact that fidelity is a matter of principle is a reason for keeping a promise even in such cases. The reasons which led me to adopt the principle, or to make the promise, may not be operative, and therefore do not in that case help to outweigh the reasons for not keeping it; but the fact that this is a matter of principle is itself a weighty reason for keeping the promise. Similarly, with other moral principles, there may be cases in which, if no principle had been involved, the balance would have come down on the other side. Arguments for rules which are supposed to be moral principles, such as the principles of toleration, must therefore show not merely that in most cases observance of the rules is the best means of furthering legitimate purposes, but that even in cases (at least, in some cases) in which that is not so the rules ought to be observed. Under some circumstances a reciprocity argument can lead to such a conclusion.
The term "toleration" has since Bayle's time fallen into disfavour, but for no good reason. One complaint is that to make toleration a virtue suggests that all sorts of evils should be tolerated.Note 5 But in this context the word means commitment to a certain set of principles: "toleration" is refraining from certain sorts of actions against certain kinds of things, the actions and things being specified by the principles.Note 6 Other kinds of action, or actions against other things, are not restricted. For example, one might argue as strongly as possible against what one regarded as religious error, one might use physical force to prevent religious fanatics from some dangerous action, and so on. Another objection is that talk of toleration implies disparagement of, and condescension toward, what one tolerates, as being an evil which it would be desirable to eliminate.Note 7 But in the present context, as a term for commitment to certain principles, there is no implication that the thing is evil, but merely that some people might think so. Even in other contexts the word does not always imply disapproval of what is to be tolerated. It is true that if I say, in the first person, that I tolerate something then that implies that I regard it as an evil. But if I call on others to "tolerate" something there is no implication that I myself regard it as an evil. When there is a tacit reference to principles, even first- person statements do not imply that the thing to be tolerated is evil or that there has ever been any sort of right to repress it. If I say I regard religious toleration (for example) as a matter of principle, that means that even if I thought that some religion were objectionable I would not act against it in certain ways. That I do actually regard some religion as objectionable is not implied.
*****
Bayle's reciprocity argument for religious toleration turns on the frightful results of wars of religion. Catholics, Lutherans and Calvinists each claim to be Jesus's only true followers and take his words "Compel them to come in" as a command to persecute the others. If Jesus was divine he must have foreseen the divisions of Christianity, and if nevertheless he commanded his followers to persecute then he is responsible for the resulting shambles. He is responsible even if he meant only his true followers to persecute, because he must have foreseen that those who mistakenly think they are his followers would take the command as addressed to themselves. But it is unthinkable that Jesus should be responsible for the horrors of the wars of religion. The words "Compel them to come in" must therefore not have been meant as a command to persecute. In fact God's law must prohibit persecution, since even without a command to persecute persecution may occur, and God must have foreseen this possibility and provided against it. The prohibition must extend even to those whose doctrines are right: otherwise it would not prevent wars of religion, since it must be applied by fallible beings who cannot distinguish being right from only thinking they are right. The rule against persecution is thus a "common principle" of morality, in the sense that it applies alike to all sects, without distinguishing those who are right from those who are wrong. This common principle makes persecution wrong intrinsically, so that it cannot be justified by its usefulness for the true religion.Note 8
Bayle's argument is for religious toleration, and it uses religious premisses (for example, that God sees the future); secular versions can be constructed for other conclusions, for example to show that rival political movements should adopt the principles of political toleration, the open society or democracy, or that nations should adopt laws of war or make treaties of disarmament, or that employers and unions should accept an arbitration system. I will use "sect"Note 9 for the parties to whom the argument is addressed, and "persecution" for the sort of behaviour it is directed against (though sometimes another term might be more appropriate - for example "atrocity", if the argument is meant to establish laws of war, or "ruthlessness", "confrontation", "rough play", etc.). Nothing is called persecution except by those who are against it; people do not claim for themselves "a right to persecute", though for brevity I will use this phrase (meaning "the right to engage in what you may call 'persecution'"). In so far as "persecution" has a definite meaning it seems to be given by rules arrived at by tacit use of the reciprocity argument (see section 5 below): "persecution" is what these rules forbid. By "persecution" I will therefore mean whatever a good reciprocity argument can be directed against, the content of the term being provided by constructing the argument.
There is not just one "reciprocity argument". Each rule, disposition or institution to be erected against a certain kind of behaviour needs a particular argument directed to a particular party. I am concerned with what such arguments have in common, with this argument as a type. The idea of reciprocity may come in twice: the argument may begin from what happens (or will happen)Note 10 when certain actions used by one sect to further its purposes are reciprocated by its rivals; and the conclusion argued for is that such actions should be renounced on condition (or in the hope) that the others will reciprocate by renouncing them also.Note 11 Thus Bayle begins with the evils which result when each of several rival sects takes the saying "Compel them to come in" as a command to persecute the others; and his conclusion is that the sects should enter into a treaty to outlaw persecution, or that they should leave the right to coerce permanently in abeyance, or acknowledge that there never was such a right.Note 12
The argument starts from a consideration of consequences -- the bad consequences of persecution, the likely good consequences of renouncing the right to persecute. But it is not a consequentialist argument; that is, it does not presuppose that the goodness or badness of an action always consists entirely in the goodness or badness of its consequences. And it is not merely an appeal to self-interest; the consequences to be considered may include effects on the progress of the true religion, and may not include effects on the worldly interests of true believers or potential converts -- what consequences are allowed to count depends on the moral code of those called on to make the calculation. The conclusion may be that some legal or other institutional barrier should be erected; or it may be that someone should adopt rules or cultivate some disposition of character to exclude some type of action as a matter of moral principle, and therefore even when such an action could be done without the bad consequences that motivated the commitment.
The reciprocity argument has no necessary connection with the idea of a social contract, but there are similarities, and the argument may lead to a contract. The classic social contract theorists start from a state of nature in which each person has certain natural rights, and argue that these rights should be renounced or qualified. Different writers postulate different natural rights: to do anything whatever, to do whatever one judges conducive to self-preservation, to enforce the law of nature, and so on. Behind these various particular rights there seems to me to be one really fundamental right, the right to act upon our own judgement of what it is morally permissible to do (or obligatory and therefore permissible). The purpose of the argument in social contract theories is to show that we should modify our view of what is permissible, or that we should renounce the right to act on our private judgement of what is permissible.Note 13 Similarly the reciprocity argument begins by postulating (or at least conceding for the moment) that there is a right to persecute (whether in the state of nature, or in a civil state in which there remains some right to act on our own judgement, or one in which the government is permitted to persecute on behalf of an established religion or other ideology), and sets out to show that that right should be abandoned, or that some institutional arrangement should be made to stop those who think they have it from exercising it. But, as I will later try to show, the renunciation of the right, although it is made in the hope that others will reciprocate, need not be conditional upon their doing so. In some circumstances the argument may lead to a contract, in others it may not.
In this essay I will explore some of the issues raised by reciprocity arguments, to decide how much they can prove and for what circumstances.
The sketchiest reciprocity argument simply invokes the saying "do unto others". To spell it out a little: relevantly similar cases should be judged the same way; so do not persecute others unless you could (or do) acknowledge that they would be entitled to persecute you if the positions were reversed. Some such argument is implicit in charges of moral inconsistency. Lutherans at first complained against persecution by Catholics, but when they were strong enough they persecuted Catholics and other Protestants.Note 14 Since they had challenged the established Catholic authorities, how could they then deny the right of others to challenge them? Since they had complained when Catholics persecuted them, how could they rightly persecute others? Their initial challenge and complaints implicitly asserted rights to think for oneself, to disagree with established authorities, to obtain a fair hearing, to preach and practise without penalty; since their later attempts at repression implicitly denied those rights they were morally inconsistent.
One possible reply is that a challenge implicitly asserts, not rights in the full sense, but only liberties, or even less.Note 15 In the minimal sense, if I have a right to do something you do wrong to blame me for doing it but otherwise you have no correlative duties. In a somewhat stronger sense, if I have a right you have a duty not to try to obstruct or deter my act intentionally; you have no duty to help, or even to get out of the way (even if you realise you are in the way), but you do wrong if you get in the way on purpose. Then there are rights in some fuller sense, which impose other duties besides the duties not to blame and not to obstruct intentionally (for example, to help). Suppose we were in a 'state of nature" (at least where ideological conflict is concerned), each of us having a "natural right" to do whatever we think right to propagate whatever we think true. Such a natural right would not impose any duty except not to blame those who exercise the right. We would have no duty not to try to prevent what we might regard as false teaching, even by force or threats (threats not of punishment, which implies blame, but of damage). We would do no wrong in doing what we thought we should do to propagate truth, or in obstructing one another's propaganda, even intentionally. There may be nothing wrong in trying to prevent something there is no blame in attempting.
Moral consistency requires that when we claim some sort of moral right we must concede that others have a right of the same sort in relevantly similar cases. What sort of right do we claim merely by speaking? A right of the lowest degree. To speak implies (perhaps) that we believe that anyone in like circumstances can rightly speak, without deserving blame, but not that anyone else has a duty to listen, or a duty not to prevent others from speaking. The use and threat of force and other disincentives are not blame and do not imply blame. Rules against the repression of speech must therefore have some other basis than the rights implicitly claimed by speaking. Those who challenge established authority may imply that what they are doing does not deserve blame, but not that there is any right to do it in a strong sense of right, and therefore not that they in their turn have any duty not to obstruct, even intentionally and by force, those who may later challenge them.
Another possible reply is that there is no inconsistency because the cases are not similar. Under one description they are similar: established authority is challenged and represses the challenge. But they are not relevantly similar. It is not the case that establishment just by itself, and nothing else, justifies authority. If positions of authority ought to be held by those with genuine knowledge and good intentions (or some other characteristic besides establishment), then a challenge to established authority by some better-qualified group is not relevantly the same as a challenge to a well-qualified authority by others who are not, and repression in one case is not the same as repression in the other.
To some this reply may seem ridiculous: the members of every sect think that they are right themselves and the others wrong.Note 16 To others it may seem shocking to suggest that truth and error may have the same rights: this sounds like scepticism or indifferentism.Note 17 But why is it ridiculous to think that knowledge is a relevant qualification for having authority? Is there some argument, not sceptical or indifferentist, to show that knowledge is not a qualification? At a given moment we cannot tell the difference between knowing and only seeming to know, but knowledge is no worse off in this respect than any other qualification. At a given moment we cannot tell whether someone who seems to have any quality really has it, though it is possible that further experience may dissipate false appearances; but it is not true universally that a quality about which we may be mistaken is morally irrelevant. We attribute rights and duties for the present to those who at present seem to have the qualities that at present seem relevant, and hope that further experience will correct some of our mistakes. If being right is to be treated as irrelevant as a qualification for authority it must be for some better reason than the possibility that what seems true may not be.
It seems, then, that the simple "do unto others" argument does not get very far. It needs to be completed by an argument to show, first, that the rival sects (parties, nations) have against one another certain rights in some sense strong enough to rule out some methods of carrying on their conflict (methods we will then call "persecution", "atrocity" etc.); and, second, that those rights and duties are not affected by the truth or falsity of their beliefs. These two points cannot be assumed, they need to be proved. That is the purpose of the more elaborate reciprocity arguments.Note 18
Bayle argues from the evils of religious warfare to moral rules against persecution, assuming that God's command is the basis of morality and that God eternally foresees these evils; he argues that God must have laid down rules to prevent them. To construct a non-theological version of the argument let us leave open the question of the basis of morality, assuming simply that the rival sects do have some morality (different moralities, perhaps), and let us set aside God's foreknowledge, and ask whether human beings can deal with the evil when they see it by applying or adapting their existing moralities and institutions. Later I will consider whether it is possible to do without any bilateral or multilateral agreement, but since current moral codes include the principle that voluntary agreements are to be kept let us explore first the possibility of a reciprocity argument leading to a contract. By virtue of the "basic" duty to keep agreements the parties would assume "additional" duties going beyond the duty not to blame or punish implied by the mere liberty to try for a hearing.Note 19 The terms of the agreement must take no account of the difference between real and apparent truth or orthodoxy (or whatever value is in dispute), not because the difference cannot be discovered or does not matter, but because it is in dispute; an agreement to reduce the evils of conflict will be ineffective if it is framed in terms which are at the centre of the conflict. Thus the two requirements which emerged at the end of section 1 will be met, without scepticism or indifferentism.
Under what circumstances would it be reasonable for rival sects to enter into a contract? I will merely indicate in general terms what sorts of things each sect's deliberations should take into account. First, the sects will have their own, possibly different, moralities, the content of which will affect the possibility of agreement. The reciprocity argument has no practical value unless it can show that even those who now believe they ought to engage in (what the other side calls) persecution should nevertheless renounce the right to do so. To take account of their position is not to endorse the belief that they have a duty to persecute. The force of the argument will be this: even if it were true that they had a duty to persecute, still in certain circumstances it would be right for them to agree not to do so in future; we may later argue that there never was such a duty, but that will be another argument.Note 20 The supposed duty to persecute might be regarded as either absolute or prima facie.Note 21 If it is regarded as absolute then that belief must be contested by some other argument before the reciprocity argument can get under way, since presumably people cannot rightly agree not to do what they believe they have an absolute duty to do.Note 22 But if they regard it as a prima facie duty they may consider reasons against carrying it out. Rules of deliberation included in their morality may exclude some sorts of reasons, or direct that they be given little weight. Perhaps the temporal cost (for example in human suffering) is not to be counted, or is to be taken into account only when other things are equal (for example, if free discussion and repressive methods are apparently equally favourable to the progress of orthodoxy, the sufferings caused by repression may be allowed to tip the balance). Perhaps the sufferings of the saints should be weighed but not those of heretics, or the sufferings of heretics should not be given as much weight; or conversely, perhaps our own suffering is not to be weighed while the sufferings of others should be; and so on. It depends on these and other aspects of the moralities of the sects concerned whether they can agree to renounce persecution without betraying or modifying their principles.
Second, they must consider relative strengths. If one sect (or coalition) is much stronger than the others, then the weaker sects have reason to offer an agreement but the stronger has no reason to accept. The agreement will not be reasonable for all unless the parties are evenly matched, or unless for some other reason the stronger cannot be sure of the outcome. If two sects are found in some other place in communication with this one and the stronger sect here is not stronger there, then both may have reason to make an agreement for both places.
Third, each must consider whether the other can be relied on to keep such promises. Some sects have believed, or have been suspected of believing, that faith is not to be kept with heretics, or that a promise must be broken if it turns out to be detrimental to God's cause.Note 23 Some sects may be unconscientious, not taking admitted obligations seriously enough. If a now powerful sect agrees not to persecute it may lose the chance to secure its position only to be persecuted by its rivals later if they do not keep their promise.
It is sometimes said that an agreement made by one generation cannot bind its successors, since we cannot be bound except by our own promise (or by a promise made by an authorised agent). Even if those who make the agreement keep their promises their successors therefore may not. There are various possible ways round this difficulty.Note 24 Perhaps the agreement can be to do something that one generation can do which will make it difficult or impossible for a future generation to persecute even if they do not believe themselves bound by agreements they have not made. For example, there might be an agreement to teach the principles of toleration as "basic" rules of morality, or to write toleration into the constitution of the country. But it may be objected that it is untruthful to teach children to regard something as a basic part of morality if you do not yourself regard it as such, and if one generation cannot make agreements on behalf of later generations then perhaps the constitution ought to be rewritten every generation.Note 25 Another possibility is to rely on the notion of succession. A successor is bound, not by the predecessor's promise, but by accepting the succession with its attached obligations. Against this it may be objected that later generations may not be successors in every respect. They may succeed the earlier generation as property-holders, for example, and be bound to pay off mortgages, but perhaps they are not bound to keep the promise to practise toleration. Another possibility is to say that they are bound to respect "vested rights", meaning rights which, because they have existed and been relied on for a long time, cannot fairly be abolished without notice or compensation.Note 26 But perhaps the rule of respect for vested rights is merely an application of the rule of respect for promises, the promise in such cases being tacit; and if a new generation is not bound by its predecessors" promises it is not bound by their tacit promises. Another possibility is to invoke Rawls's suggested duty of fair play.Note 27 But this applies only to those who have voluntarily accepted benefits, and a later generation may claim that it had no choice, since toleration was already the practice.
Whether the other parties to an agreement believe -- rather, whether their successors believe -- that later generations must keep agreements made by earlier generations determines whether the agreement will be enforced by the consciences of future generations of members of the agreeing sects. But other people's beliefs are also relevant. They might be outraged (though for no reason we have been able to discover) if the agreement were repudiated for the reason that an earlier generation cannot bind its successors, and they might help enforce it out of a (possibly mistaken) sense of interest in an unrestricted version of the principle that agreements are to be kept.
Even if it is doubtful that the other party can be trusted or that later generations will feel bound, it may still be reasonable to make the agreement, gambling on the likelihood of some change of belief when conflict diminishes, or on the stability of peaceful habits. If the practice of toleration becomes well established, the question whether it should continue may not occur to the new generation of a sect even if the balance of power has shifted their way.
There are, then, circumstances in which an agreement to outlaw persecution might not be reasonable: the rival sect might be weak and likely to remain so, one's own moral code might make what others will call persecution an absolute duty, the other party might not be trustworthy, or it may be possible that their successors will not feel bound. In such cases a reciprocity argument might not work, at least not without preliminary argument to change some of the parties" moral or factual beliefs.
Two or more sects which agree not to persecute one another may still persecute others. If one of the others turns out to be well able to reciprocate it may also be admitted to the agreement, but a non-persecuting sect might not have enough bargaining power to get admitted. What reason might there be to agree not to persecute a non-persecuting sect? The fear of being persecuted is not the only possible inducement to agreement; a milder inducement, which may sometimes be enough, is dislike of being subjected to precautions. There is an important difference between persecuting the intolerant and taking precautions against them. But even if the precautions are really precautions and not disguised persecution, and even if they are carried no further than reasonable fears justify, they may bear quite heavily on members of a feared sect.Note 28 If a powerful but non- persecuting sect takes precautions against sects which might persecute it, this may be reason enough for the latter to agree with it to outlaw persecution.
Suppose now that the terms of the agreement are altered, and the parties promise not merely not to persecute one another but not to persecute anyone, including those who do not join in the agreement and those who join but later violate it, but reserving the right to take reasonable precautions against non-joiners, and against those whose beliefs or behaviour suggest that their adherence may not be sincere. To frame the agreement this way may make it more difficult for the reciprocity argument to succeed. In taking "reasonable precautions" instead of persecuting there is a sacrifice of security which a sect now powerful enough to persecute may regret if others grow powerful and do persecute. Sects which do not join at the beginning are less likely to join later, since being subjected to precautions is not as strong an inducement as being subjected to persecution. But these drawbacks may be balanced or overbalanced by the fact that an agreement in these terms is likely to be more stable and more effective. Acts of persecution by unrepresentative members of a sect, or a false rumour of persecution, or a real but perhaps temporary relapse, will not be so likely to cause the collapse of the agreement and the renewal of outright warfare; it will lead only to a heightening of precautions, providing an incentive to the erring party to honour the agreement in future. Also, the practice of toleration uniformly toward all, rather than only toward those who join in an agreement, is more likely to develop into an unquestioning habit, which, as was noticed above, may be the best solution to the problem of making the agreement last through generations. As long as some sects are persecuted it is obvious that the toleration of others rests upon voluntary agreement, and the practice of toleration cannot become unquestioning habit.
References to inducements and security do not mean that the calculation is self-interested. It is to be done in accordance with the deliberator's moral code, which will determine the extent to which selfish interests may be taken into account. If all that is allowed to matter is some ideal cause, then security in the relevant sense has been sacrificed if the cause is jeopardized, and the inducement is benefit to the cause. The worldly interests of individual sect members may not count at all.
So far we have considered two kinds of reciprocation. First (in section 2) we imagined an agreement by which one sect promises not to persecute the other, on condition that the other makes the same promise, while each keeps the right to persecute those who are not parties to the agreement. Then (in section 3) we imagined another agreement in which each promises not to persecute anyone at all, on condition that the other makes the same promise, while each keeps the right to take reasonable precautions. Now we come to a third kind, in which a sect or individual, without waiting for agreement with others and without making the promise conditional upon anyone else's promise, undertakes not to persecute anyone, reserving the right to take precautions. The undertaking is unilateral and unconditional, but it looks for some reciprocation.Note 29 It is made in the hope that others will follow suit, now that they feel less threatened by persecution and can expect some relaxation of precautions against them if they also renounce persecution.
Just as in the second version there is a sacrifice of security in comparison with the first, so in the third there is a sacrifice in comparison with the second: those who renounce persecution without waiting for others may lose a chance to strengthen their position without getting anything in return, since after all the others may not reciprocate. Thus there may be some circumstances in which an unconditional and unilateral promise would not be sensible, in which it would be better to seek an agreement. But just as the second version has an advantage which may overbalance the risk, namely that it is more stable (see above, sect. 3), so has the third, namely that it is easier to bring about. When there are many sects their suspicions may make it difficult to organise an agreement that includes enough of them to secure peace; a unilateral declaration may then be as much as any sect can do toward peace, and may be worth the risk. An agreement as in the second version between some sects is a unilateral declaration on their part in relation to the others. A movement toward toleration may begin with a unilateral declaration by one sect, or an agreement of the second kind between sects, and be carried forward by unilateral commitments from other sects who wish to avoid being objects of precaution. In fact the movement may be begun and carried forward by individuals. An agreement between sects must be made by representatives able to speak for the sects, and leaders are often unable or unwilling to make new and risky commitments. But unilateral declarations can be made by individuals speaking for themselves. Anyone can announce that he or she will no longer behave in certain ways (and refer to them as "persecution"), and call on others to renounce such methods also.Note 30 The practice of toleration can spread without waiting for sect leaders to be convinced, behind their backs or over their heads.
Thus a reciprocity argument need not be an argument for a contract. Under some circumstances it may favour a unilateral commitment. A commitment need not be a promise.Note 31 In general terms, the argument is meant to show that some sort of guarantee should be given that certain kinds of things will not be done. A guarantee may consist in physical things (as when the French Protestants were allowed under the Edict of Nantes to fortify certain places), or in social arrangements (for example, it may be established by a public promise attaching the sanction of public opinion, or a constitutional amendment attaching legal sanctions). It may consist in a disposition of character, in being the sort of person who can be relied on not to act in certain ways, whose conscience rejects and disapproves such actions; so we might cultivate in ourselves or in others (for example children) a commitment to a certain rule or ideal as a "basic" duty of morality, or by means of a promise make it an "additional" duty of morality. A reciprocity argument is an attempt to show that some sort of guarantee should be given on condition that others reciprocate, or in the hope that they will, as a way of ending or reducing the evils which result when several parties exercise or carry out some alleged duty, right or liberty against one another. The argument is that whatever ideal and other ends the moral code requires or allows are best served on the whole if certain actions which might sometimes have been useful toward those ends are renounced as a matter of principle, or made impossible or difficult or costly.Note 32
If the guarantee consists in accepting something as a moral duty, then persecution (aggression, atrocity, undemocratic behaviour, or whatever the actions outlawed are to be called) is morally wrong even when it is expedient: toleration (for example) has become a matter of moral principle.Note 33 Many of the classic arguments for liberal principles are not strong enough to reach this conclusion. For example, Mill's argument in Chapter 2 of On Liberty to show that truth is best served by freedom of discussion does not show that an occasional judicious act of censorship (for the sake of truth, or for some other legitimate end) must be morally wrong. The general usefulness of a policy does not make it morally wrong to depart from it even when a departure would be useful. But if a reciprocity argument shows that we should commit ourselves, as a matter of principle, not to practise censorship (for example), and we do make the commitment, then any act of censorship is morally wrong, by virtue of the principle that commitments should be kept. But this is not always the outcome of the argument. There are various possible guarantees besides morality, and various possible ways of setting up a guarantee (for example, contract or unilateral commitment); in some circumstances a reciprocity argument might succeed for some of these possibilities and not for others. There may be some circumstances in which no reciprocity argument works at all.
Historians sometimes suggest that religious toleration came about first as a "peace of exhaustion", by a process in which moral principle played little part: the warring sects gave up persecuting when they found that they were making little or no headway at great cost. It is suggested that it was only in later times that toleration and other liberal practices became genuinely a matter of moral principle for all except some who would still persecute if they could. But perhaps moral reasoning played a larger part. Reciprocity arguments might under changing circumstances justify a series of steps such as in fact were taken: religious war giving way to "non-aggression pacts" between neighbouring sects; these giving way to general toleration, as compacts give way to unilateral and unconditional commitments to tolerate all who are not intolerant, and as more sects and individuals see the advantage of such commitment (the advantage increasing as the party of toleration increases); persecution of the intolerant giving way, especially when the intolerant have been outnumbered, to toleration of all with reasonable precautions against the intolerant; and deliberate adoption and proclamation of tolerant principles giving way to unquestioning acceptance of toleration as a matter of "basic" morality. In the end, when toleration has been taught as part of morality in many families for several generations, the intolerance of those who lived before the movement toward toleration began will seem immoral and the calculating and bargaining unprincipled. But the process may well have been motivated and guided by reasoning of the sort I have been analysing, starting from a moral code, deliberating in ways permitted by it, and arriving at a revised morality (which may permit other modes of deliberation, leading to further revision, and so on).
The outcome is a complex set of rules regulating conflict arising from religious beliefs, and from beliefs connected with or like religious beliefs, such as moral and political convictions. Some of the rules assert rights or duties -- some absolute, others imperfect or prima facie -- and others are rules of deliberation which limit what can be weighed against these rights and duties. The rights and duties run more or less as follows. Concerning thought: people are not to be punished for what they believe, or for changing their beliefs;Note 34 negligence or self-deception or other faults are not to be inferred merely from what they believe;Note 35 we all have a prima facie right to investigate any question as we see fit;Note 36 there is to be no compulsory religious instruction.Note 37 Concerning expression: we all have a prima facie right to express and advocate our beliefs to any audience and by our own methods, and a right not to do what may seem like an expression of belief (an absolute right, if the act is normally taken as an expression of religious belief);Note 38 we have a duty of imperfect obligation to give a serious hearing to the beliefs of others. Concerning action: we have a prima facie right to act on our beliefs, and a prima facie duty not to tempt others to act against conscience;Note 39 we have an absolute right not to perform an act normally taken as religious. Concerning public neutrality toward religious and similar beliefs: we all have a duty of imperfect obligation to make it easy for other people to take part in community life, whatever sect they belong to;Note 40 minority beliefs should get a fair share of attention in public discussion; members of organisations open to the public have the right to exercise a veto, or to contract out, when the organisation is about to do something against their beliefs; the State and public opinion are not to enforce religious or moral rules as such; public institutions are not to profess or help (intentionally and directly, at least) any religion or similar doctrine;Note 41 if we make any unfavourable moral judgements of others we must judge them according to their lights;Note 42 in taking precautions we must not give much weight to the presumption that the members of a sect all think and act alike.Note 43 The rules which assert prima facie or imperfect rights or duties put a weight into the balance, a heavier weight the higher the ideal of an open society is supposed to come in the scale of values. The rules of deliberation which limit what can be weighed on the other side seem to me to be best stated as excluding certain kinds of reasons, rather than as listing what can be given weight.Note 44 Rules of deliberation: Against the rights and duties set out above no weight is to be given to any allegation that some belief is bad in itself, or unlike the beliefs of others (against "community standards"), or that it has practical implications contrary to religion, law,Note 45 morality or public opinion. The details of this account of liberal social principles may be disputed, but the point I want to make is that the rules are complex and rather untidy. This is more easily explained if we assume that they are like clauses of a peace treaty worked out under the guidance of reciprocity arguments, reflecting the intensity of various kinds of conflicts and the distribution of power in post-Reformation Europe. The rules of deliberation against giving weight to certain kinds of reasons are directed against the motives which have led in the past to the worst conflict. The absoluteness of rights and duties relating most directly to religion reflects the length and bitterness of the religious struggle. Since so far (perhaps because dissenters have always been outnumbered) there have been no "wars of morality", except in so far as morality is involved in religion, the rules relating to morality are not absolute. Thus the State is permitted to enforce a socially beneficial rule which is part of morality provided it is not enforced because it is part of morality, whereas the State cannot enforce participation in religious worship or support specifically religious education even when that would improve social cohesion or serve some other legitimate secular purpose. Such features seem easiest to explain on the hypothesis that the rules were arrived at through reciprocity arguments. None of the other arguments for toleration leads to such a complex conclusion.
Most people these days seem to regard the principles of toleration as permanently and universally valid moral principles; the intolerance of past ages may be excused by ignorance, but objectively it was wrong. But reciprocity arguments will not justify this judgement. During the Middle Ages in Europe Catholics were much stronger than other sects and had no reason to doubt that they always would be. Communication between Europe and other parts of the world was poor, so missionaries could be sent to distant lands without expecting that they would be excluded or persecuted because of Catholic intolerance in Europe. Given these facts and their beliefs, medieval Catholics could not reasonably have thought it best to commit themselves to toleration as a matter of principle. The reciprocity argument for toleration would not have worked if it had been addressed to them.
But although it would probably not have convinced them, perhaps objectively it would have been a good argument, since it was always possible that the dominance of the Catholic Church would come to an end. Medieval Catholics should have taken account of that possibility, and should have provided for it by adopting the principles of toleration. And it might be argued that, in general, history shows that no group can be sure it will never need the protection of the principles of toleration, so everywhere and always it is wise for everyone to acknowledge these principles. The answer to these arguments is that they make a mistaken assumption about what caution requires. A group (a nation, for example) would not increase but reduce its security if it now committed itself to the peace treaty it might have to accept if it lost its present advantage, no matter how unlikely it is that this will happen soon. It would not have been reasonable for the medieval Church to commit itself to the principles of toleration while its dominance seemed secure. The force of the reciprocity argument is limited. The most it can show, I believe, is that it is reasonable to adopt the rules of toleration when the circumstances are as they have been, and are likely to continue to be, in modern times. From now on they can be treated as part of "basic" morality, but at the risk of distorting our judgement of the morality and intelligence of people in the past.
John Rawls's theory might seem to provide a way of escaping this limitation. On his account a social contract is agreed upon (the agreement must be unanimous)Note 46 in ignorance of the relative strength of sects and in accordance with an extreme view of what caution requires. He believes that the contract would provide for equal liberty of conscience, limited only by the interest in public order and security. In the Original Position the contractors do not know whether in ordinary life they belong to a strong sect or a weak one, and do not know what in ordinary life their moral and religious convictions may be. But they know that they may have moral and religious convictions that impose absolute obligations. In deciding the rules of justice they assume the worst -- that an enemy will assign their place, as Rawls sometimes puts it; so they assume that they will be in a weak position. To make sure that, in a weak position, they will be able to carry out whatever absolute obligations they may believe they have, they must adopt rules which guarantee equal liberty of conscience. But they also agree that the State may regulate liberty of conscience for the sake of public order and security, provided the regulations can be shown to be necessary by "ordinary" modes of argument. The State has no moral or religious function; its duty is limited to underwriting the conditions of equal moral and religious liberty, which it does by ensuring public order and security.Note 47
But even apart from questions about why we should go behind such a veil of ignorance, and about the reasonableness of the cautious decision rule, this argument does not seem to work. It is possible (and presumably the possibility will be known behind the veil of ignorance) that the moral and religious convictions some people have in ordinary life oblige them to persecute. If the contractors must hold themselves free to act on their possible moral and religious convictions, then they must hold themselves free to persecute. Rawls answers that, even so, the principle of equal liberty of conscience must be agreed upon if any principle can be.Note 48 But perhaps none can be. If the deliberators behind the veil of ignorance must reserve enough freedom to discharge absolute moral and religious obligations, and if these may include obligations to persecute, and if they must follow a cautious decision rule, then they cannot agree upon any principles. They must leave ideological conflict unregulated by rules of right and duty. They will not adopt rules of toleration, or any rules or institutions which might hinder intolerant acts which some may think they have an absolute moral or religious obligation to perform. There will therefore be no State to underwrite the conditions of equal liberty of conscience by preserving order and security, since the State might stop some from carrying out what they may believe is their duty to persecute. If there is a State, it must be prohibited altogether from hindering conscientious acts, including acts of persecution.Note 49
If the "veil of ignorance" lifts partly to allow the original contractors to discover whether in ordinary life they believe they ought to persecute, then some will discover that they do. Unanimous agreement on any rules or institutions will still be impossible unless they are allowed to discover the relative strengths of the parties, and it may be impossible even then. But once the veil of ignorance has lifted so far, the argument is a reciprocity argument of the sort I have been discussing,Note 50 and has the same limitation.
The old controversies about religious toleration are paralleled today by disagreements over the rules of liberal democracy. These rules impose certain moral obligations: to obey the law until it is changed by the proper parliamentary process, in extra- parliamentary political action to respect the law, to respect the equal political rights of other people, and so on. Some critics of the existing order openly reject these rules. Others say they accept them, but encounter contempt from one side and distrust from the other, just as liberal Catholics did at first. It is suspected that they have abandoned the fight, worn down by the moral pressure of liberal precautions against them and seduced by the prospect of respectability, or that their apparent conversion is just a trick.
As has often been pointed out, the rules of liberal democracy work to the advantage of those who already get most out of the existing social order. The parliamentary process is generally biased against parties of the left, and so are the laws which extra-parliamentary action is supposed to respect. Parliamentary government is in fact one part of a dual regime, the other part consisting in the private oligarchical governments which rule most of the everyday life of most people, namely the managements of business firms, schools and universities. Since private business controls a large part of economic activity, government needs business confidence. Equal political rights for all groups are most beneficial to those who have money and time, since they can exercise their rights more effectively. Everyday life conditions people to take for granted ideas favourable to the existing state of things, so when the critics get a hearing they sound foolish. Even if they are persistent they make no headway. The oligarchs need give no reasons in defence of their power, they only have to sit tight and wait for the critics to give up or die.
Why then should those who believe that the existing state of society is unjust or exploitative accept the rules which regulate ideological competition, and accept them as being a matter of moral principle? To be acceptable to all parties, including the stronger party, rules designed to moderate conflict must reflect the existing distribution of power. The only alternative is to have no rules. Rules supported by a reciprocity argument will therefore on many occasions give certain advantages to the stronger party. Even so, to guarantee obedience to the rules may on balance and in the long run be better for the weaker party or for the cause they represent than the alternative, having no rules. Under some circumstances, therefore, it may make good sense for opposition and other minority groups to make a sincere commitment to biased rules.
Under some circumstances this may not be true. Reciprocity arguments do not always deserve to succeed. Just as those brought up to regard the rules of religious toleration as part of basic morality often cannot see the calculating and bargaining that took place in the past as anything but immoral and unprincipled, so well-brought-up liberals often find it hard to sympathise with those who will not commit themselves to the rules of the open society. But in so far as the commitment rests on reciprocity arguments, people have a right not to commit themselves to obedience if they think that this would, without a good enough reason, stop them from doing things they believe they ought to do; and if they do commit themselves and then come to feel that that was a mistake they should be able to withdraw, with due notice and perhaps with some compensation to those affected by their withdrawal. But of course other people have the right to take precautions against those who refuse or withdraw the commitment, however reasonable (by their lights) that may be.
Rawls has made several references to reciprocity, but not in my sense. In one place he says that what makes reciprocity "fundamental to both justice and fairness" is the "requirement of the possibility of mutual acknowledgment of principles by free and equal persons who have no authority over one another".Note 51 Elsewhere he calls reciprocity the propensity in human beings to respond in the same way to kind and fair treatment.Note 52 The first reciprocity might be compared with the thought that figures in reciprocity arguments, that if we persecute they will too, since they have as much right. The second reciprocity might be a reason for hoping that if we adopt the principles of toleration they may too.
Barry's "Justice as Reciprocity", sects. IV-VI, is concerned with something like what I call reciprocity arguments, and draws attention to limitations like the ones I point out. He suggests that reciprocity is not a strong argument for foreign aid, since rich countries are not likely to need aid from poor countries. David Richards, also concerned with foreign aid, suggests that the limitations of reciprocity can be overcome if we substitute what he calls "moral reciprocity" for "actual reciprocity", moral reciprocity being the idea of "treating persons in the way one would oneself reasonably like to be treated".Note 53 This looks like the golden rule, discussed above in sect. 1. The only way I can see to define how it would be reasonable to like to be treated, and what is to count as treating people that way, is to construct a reciprocity argument, which is concerned with actual reciprocation (see above, n. 10), and the same limitations would be encountered.
Part of the job in constructing a reciprocity argument is to find the right level of abstraction (see above, n. 18). Individual citizens of a rich country may have reason to commit themselves to the rule that anyone, anywhere, at any time, who needs help to stay alive should be helped (or something similar), and this implies a duty to support foreign aid. More likely they will have reason to commit themselves to some more restricted rule, for example that people in need should be helped by relatives, friends or neighbours; or to a more complicated rule specifying an order in which others are to help -- if the neighbours cannot help, then. . . . Residents of a rich country might be obliged by this rule to help residents of poor countries when closer help has failed.
Return to Home Page
Note 1. Cf. Mill on the "transcendant expediency" of regarding veracity as a matter of principle, Utilitarianism, ch. 2 (p. 223).
Note 2. In a letter Arnauld remarks that if even half of what is reported about the coercion of the Huguenots is true it is deplorable, and likely to make the Catholic religion odious. But what he deplores is that coercion is being used without adequate provision for instruction. See Oeuvres, vol. 2, p. 336 (misprint for p. 136). In his Apology for the Catholics he adopts St Augustine's opinion that coercion may be justified if it makes people willing to listen and examine (Oeuvres, vol. 14, pp. 717ff.). Cf. Lecler, vol. 1, pp. 56-7.
Note 3. Cf. Ross, pp. 19ff. There are also duties of "broad" or "imperfect" obligation (e. g. the duty to be generous), which do not require any particular act at any particular time, but require us to do "a certain amount" over time. Cf. Mill, Utilitarianism, ch. 5 (p. 247), and Kant, Virtue, pp. 37-8, 48-9. On a particular occasion one may decide that one ought to do something generous, but this "ought" does not signify strict duty.
Note 4. The "weight" due to a consideration of a certain general kind seems to be different in different cases; it is the consideration as realised in this case which has to be weighed. This is why intuitive "weighing" cannot be replaced by deductive application of rules. See Ross, pp. 41-2, and Mill, Utilitarianism, p. 259 ("though particular cases may occur in which some other social duty is so important as to overrule any one of the general maxims of justice"). It would seem that considerations of moral duty do not in every case outweigh considerations of other kinds; sometimes a less serious moral obligation may be set aside for the sake of some legitimate but not obligatory purpose. See Foot, pp. 181ff.
Note 6. Cf. Benn, "Privacy", pp. 1-3, on privacy as a "norm-invoking" concept.
Note 7. Kant praises the prince who leaves people free to use their own reason in matters of conscience but "declines to accept the presumptuous title of tolerant" (Enlightenment, p.58). Tom Paine (p. 107) says that toleration is a despotism which assumes to itself the right of granting liberty of conscience.
Note 8. See above, Essay II, sect. 5.1.
Note 9. In using "sect" I do not intend any contrast with "church".
Note 10. The reciprocity argument is not concerned with hypothetical universalisation ("What if everyone did this?"), but with the likely consequences of one's own actions under existing circumstances.
Note 11. It is the second reciprocity that I regard as characteristic of this kind of argument. Perhaps the others will persecute us whether we persecute them or not, but if we renounce persecution as a matter of principle (which is more than simply not practising it), perhaps they will follow suit. The argument is about the likely consequences of committing oneself to a principle when the consequences may include a similar commitment by others.
Note 12. See above, Essay II, sect. 5.1.
Note 13. See my article, "Locke on Political Obligation".
Note 14. See Lecler, vol. 1, pp. 147ff., 242ff.
Note 15. See Hart, "Natural Rights", pp. 56-7. Hart says that a liberty imposes no duties on others. It seems to me that every liberty imposes a duty not to blame, and some liberties (or should we call these rights?) may impose a duty not to obstruct intentionally. Perhaps the duty not to obstruct is imposed by some part of what Hart calls "the protective perimeter" ("Bentham", pp. 179ff.); but, even so, to postulate a liberty implies some duty, by bringing into play some protective principle.
Note 16. "If he says that he is to be spared because he believes true, but the other was justly persecuted because he was in error, he is ridiculous. For he is as confidently believed to be a heretic as he believes his adversary such" (Jeremy Taylor, Liberty of Prophesying, p. 517). Cf. the quotations at the beginning of Dworkin's article.
Note 17. "Right is a moral power which. . . it is absurd to suppose that nature has accorded indifferently to truth and falsehood"; "they. . . end at last by making no apparent distinction between truth and error. . . [T]he Church. . . is forced utterly to reprobate and condemn tolerance of such an abandoned and criminal character" (Pope Leo XIII, in Gilson (ed.), The Church Speaks to the Modern World, pp. 72, 78).
Note 18. To "treat like cases alike" we must know how to decide which cases and which "treatments" are alike; see Hospers, pp. 286ff. Substantive rules, such as the rules of toleration, help determine which similarities and differences are to be regarded as equivalent. In deciding whether to commit ourselves to such rules we must consider the consequences of "treating like cases alike" according to the classification implicit in the rules. Which cases and treatments are to count as alike is decided by deciding on the rules.
Note 19. By a "basic" duty I mean one directly prescribed by the moral code (even if it could be derived from other rules, it is directly prescribed). By an "additional" duty I mean one which arises from some action by virtue of a basic duty, e.g. the duty to do something arising from the act of promising to do it by virtue of the basic duty to keep promises, or a duty arising from a superior's command by virtue of the basic duty (if there is one) to obey the powers that be. In different moral codes additional duties may arise in different ways.
Note 20. A reciprocity argument may thus have some impact on people who subscribe to "non-neutral" principles. "Compel them to come in", understood as addressed only to the true religion, is a classic "non-neutral" principle, one whose "application to particular cases is a matter of controversy for the parties whose conduct is supposed to be regulated" (Dworkin, p. 492). Bayle's argument seeks to establish "common principles" which can be applied to regulate conflicts without the need first to decide the questions in dispute (see above, Essay II, n. 163).
Note 21. On this terminology see above, Introduction.
Note 22. Very likely they will regard the absolute duty as something intuitively obvious. We might suggest some other formulation and ask them to consider whether it is not rather this that seems to be their duty. Or we might try to invent some analogous case in which their principle loses its intuitive appeal. Or there might be some other principle that could be brought in some cases into conflict with this one. "When an impudent rascal blasphemes. . . the authority who allows such an act and does not punish it severely shares before God in that sin" (Luther, quoted Lecler, vol. 1, p. 156). Could the principle rather be that we must dissociate ourselves in some way (e.g. by protesting), without necessarily inflicting punishment? What about the analogous case of other sins: to avoid sharing in sin must the ruler search out rigorously every sin of whatever sort and punish it?
Note 23. Catholics were suspected of believing this after John Hus was burned despite a safe conduct. Some Protestants adopted the same maxim. Suspicion over this was an obstacle to the development of toleration. See Lecler, vol. 1, pp. 290-1, 302; and vol. 2, pp. 236-7, 241-2.
Note 24. This is not the same as the difficulty about reciprocity between generations (see Barry, "Justice as Reciprocity", pp. 69ff.). The agreement I am concerned with would regulate relations between contemporaries; the problem is whether it will be reproduced into the future. "As the ends of such a partnership [the State] cannot be obtained in many generations, it becomes a partnership not only between those who are living, but between those who are living, those who are dead, and those who are to be born" (Burke, p. 106). This is not right: there is no partnership between members of different generations. Rather, since the end sought by members of this generation cannot be attained without an agreement which will not be made unless it is expected to last beyond one generation, there must be some assurance that members of new generations will join in the same agreement with other members of their own generation.
Note 25. "Each generation. . . has. . . a right to chose for itself the form of government it believes most promotive of its own happiness;. . . and it is for the peace and good of mankind, that a solemn opportunity of doing this every nineteen or twenty years, should be provided by the Constitution" (Jefferson, p. 675). But it is unlikely that a federation could ever be formed unless the parties believed that the constitution would last more than one generation. On binding future generations see Locke, Second Treatise, sect. 116; Rousseau, pp. 169-70; Bentham, pp. 494, 501; Paine, pp. 63-4. If (as Hume argues) the practice of promising is based on a sense of its utility, the utility of being able to make arrangements that will last beyond one generation is a reason for rejecting the view that promises cannot bind successors.
Note 26. See Barry, Political Argument, p. 102. "Prescription" is a similar notion. See also Hume, Treatise, vol. 2, pp. 254-5.
Note 27. On fair play see Rawls, Justice, pp. 111-12.
Note 28. On precautions see above, Essay II, sect. 6.3. The exclusion of Catholics and Dissenters from public employment in England until the nineteenth century was defended by some as a reasonable precaution and criticised by others as virtual persecution. See Cobbett, vol. 9, cols. 1047-59, vol. 26, cols. 780-831, and vol. 28, cols. 387-452; also Hansard, n.s. vol. 4 (1821), cols. 949-1030; Hoadly, passim; Paley, pp. 472-7; Macaulay, vol. 1, pp. 7-9, and vol. 2, pp. 255ff.; Mill, Prefaces, p. 151. A similar debate took place in the 1950s about similar precautions against Communists; see Hook, and articles by Hook, Lovejoy and Lowe in Journal of Philosophy 48 (1951), and 49 (1952). Precautions may take other forms, such as preventive detention and surveillance.
Note 29. This is common behaviour. Acquaintances often do one another favours partly in hope of return, but it would be unfriendly and self-defeating to attach conditions in every case, or even to be too conscious of the balance of the account.
Note 30. Or, to take a topical instance, we could individually say that we will have nothing to do with nuclear weapons or the materials from which they can be made, no matter what others may do.
Note 31. "When this common sense of interest is mutually expressed, and is known to both, it produces a suitable resolution and behaviour. And this may properly enough be called a convention or agreement betwixt us, though without the interposition of a promise"; "Every member of society is sensible of this interest: every one expresses this sense to his fellows, along with the resolution he has taken of squaring his actions by it, on condition that others will do the same"; Hume, Treatise, vol. 2, pp. 195, 202. As the second passage indicates, mere expression of a sense of interest is not enough; what is needed is the expression of a resolution, indeed an undertaking or commitment, though it need not take the form of a promise. And it need not be "on condition".
Note 32. See my paper "Utilitarianism and Virtue". The reciprocity argument is a subclass of the kind of argument explored there.
Note 33. When liberal Catholics began to favour toleration liberals suspected a strategem, that if the Church ever became strong enough to persecute again it would do so. Archbishop Dupanloup's distinction between thesis and hypothesis seemed to mean that Catholics regarded toleration as right only for the prevailing circumstances; see Vidler, p. 152. "Eurocommunism" has met with similar suspicion. To count as genuine, the conversion must be to the belief that toleration is a matter of principle.
Note 34. Hardly anyone has thought that people can rightly be punished just for misbelief, but many have held that those who have once accepted the true faith can be punished if they abandon it; see e.g. Thomas Aquinas, Summa, 2-2, q.10 a.8.
Note 35. Some have thought that there are errors no one can fall into (e.g. about basic moral principles) without being guilty of negligence or self-deception or some other fault. The principle suggested in the text requires that there be specific evidence of such faults; they are not to be inferred simply from what the person believes. See above, Essay II, sect. 4.2, and Locke, Third Letter, pp. 297-9.
Note 36. Christians have often believed that it would be wrong for them to "check up" on the truth of their religion. Conservatives have sometimes taken the view that it is blameworthy for ordinary people to pry into the foundations of society. (According to the Bishop of Rochester the Treasonable Practices Bill of 1795 was "merely directed against those idle and seditious public meetings for the discussion of the laws where the people were not competent to decide upon them. In fact, he did not know what the mass of the people in any country had to do with the laws but to obey them, with the reserve of their undoubted right to petition against any particular law as a grievance on a particular description of people" (Cobbett, vol. 32, col. 258).) Sometimes it is believed that certain inquiries are permissible only if carried on in certain ways (e.g. by asking respectful questions of the authorities). To act on such beliefs oneself, or to persuade others to act on them, is compatible with commitment to toleration (to be tolerant it is not essential to be a free-thinker), but not to enforce them upon others.
Note 37. Melanchthon defended compulsory attendance at sermons as compatible with liberty of conscience: "The Prince does not force the spirit, but only the motory faculty [i.e. he makes sure they do not walk out]; he forces people to listen to the true doctrine, and he forbids outward blasphemy; after that, the listener keeps the duty and the power to know the truth, if he wants to" (quoted Lecler, vol. 1, p. 253).
Note 38. On the right not to express a belief (even if you actually hold it) see Konvitz, Liberties, pp. 111-16.
Note 39. On temptation see above, Essay II, sect. 5.2.
Note 40. Ordinary good manners require considerateness toward members of minorities -- to acknowledge their existence, to provide food they can eat, to avoid words they regard as offensive, to take account of their religious holidays and so on. Toleration is concerned with such things in so far as minorities are constituted by their beliefs. As requirements of toleration they may have more weight than they have as requirements of good manners.
Note 41. The State subsidises religion indirectly if it subsidises private hospitals and some of them are religious. The subsidy to religion might be intentional: the policy might have been designed to provide certain religious groups with income. Or it may be unintentional (even if the benefit to religion is known): the purpose might be to extend hospital services. On controversies involving these distinctions see Konvitz, First Amendment, pp. 12, 16, 23-6, 60.
Note 42. See above, Essay II, sect. 3. This applies to unfavourable judgement only. Toleration requires that we respect those who live by their own code, not that we judge them unfavourably if they do not. In any case, their code may be excessively demanding.
Note 43. The difficulty of predicting what people will do from what they believe was one of Bayle's favourite themes; see Miscellaneous Reflections, sects. 135-6, 176, 181. Cf. Fox, vol. 4, pp. 2-4; Macaulay, vol. 1, pp. 7-9. This was one of Macaulay's arguments against excluding Catholics from government jobs; see above, n. 28.
Note 44. For an example of the method of listing what can be weighed, see McCloskey, p. 221; for an example of the method of listing what is not to be weighed see Scanlon, pp. 209, 213. The advantage of the latter is that it does not need a complete list of possible reasons for action. The point of the principles of the open society is to exclude certain kinds of reasons for interfering with others. Toleration should not be formulated so as to tie it closely to any particular system of positive values, since people who hold different systems are all supposed to be tolerant. (Mill's principle of liberty as I understand it -- see my article "Mill on Duty and Liberty" -- says what can be a reason for interference, namely some violation of duty to others; but in deciding what to regard as duties to others we are governed by the rules of deliberation stated above, which say what cannot be a reason.)
Note 45. I do not myself acknowledge any moral duty to obey the law because it is the law, or because it has majority support. Those who think there is such a duty will think that if the act would be illegal, that counts against the prima facie right to act on one's beliefs. But it does not count against the other prima facie rights and duties. For example, advocacy (short of incitement) is not to be judged wrong because the belief advocated is that some illegal act ought to be done; people must be free to discuss whether they ought to obey a law.
Note 46. In reality no contract is involved; see Snare and Hampton. The theory is that the rules of justice are those which it would be rational for any person to adopt if he or she were ignorant of (or disregarded) certain facts and had to choose the same rules for self and others. Whether there is any actual consultation and agreement makes no difference to what those rules would be.
Note 47. See Rawls, Justice, pp. 205-21. Restrictions on liberty of conscience must be justified by modes of reasoning which are "acceptable to all", "ordinary", "not controversial" (p. 213). There may be none; and the exclusion of some modes of reasoning (e.g. from the bible) may be controversial. The reciprocity argument begins from the existing convictions and ideas about reasoning of each of the parties to whom it is addressed, and (if it succeeds) arrives at rules which may include rules of deliberation limiting the reasons which may be appealed to in certain situations, permitting Rawls's "ordinary" modes of reasoning (and any peculiar modes not ruled out). What sort of reasons are to be regarded as acceptable is thus decided by the outcome of the argument, not assumed without argument by appeal to a consensus. See above, n. 18.
Note 48. Rawls, Justice, p. 208.
Note 49. Such a prohibition would be a considerable restriction of state power. Contrast Bayle and Locke, above, Essay II, sect.6
Note 50. Not quite. Rawls's argument is about a hypothetical decision (the rules of justice are the rules it would be rational to adopt under certain conditions), whereas a reciprocity argument is a reason for establishing a guarantee, and requires completion by an actual decision and action; toleration becomes a moral duty only when a commitment is actually made.
Note 51. Rawls, "Justice as Reciprocity", p. 256.
Note 52. Rawls, Justice, pp. 496ff.
Return to Home
Page