§4. I pass to consider another group of duties, often contrasted with those of Benevolence, under the comprehensive notion of Justice.
``That Justice is useful to society'', says Hume, ``it would be a superfluous undertaking to prove'': what he endeavours to show at some length is ``that public utility is the sole origin of Justice'': and the same question of origin has occupied the chief attention of J. S. Mill. Here, however, we are not so much concerned with the growth of the sentiment of Justice from experiences of utility, as with the Utilitarian basis of the mature notion; while at the same time if the analysis previously given be correct, the Justice that is commonly demanded and inculcated is something more complex than these writers have recognised. What Hume (e.g.) means by Justice is rather what I should call Order, understood in its widest sense: the observance of the actual system of rules, whether strictly legal or customary, which bind together the different members of any society into an organic whole, checking malevolent or otherwise injurious impulses, distributing the different objects of men's clashing desires, and exacting such positive services, customary or contractual, as are commonly recognised as matters of debt. And though there have rarely been wanting plausible empirical arguments for the revolutionary paradox quoted by Plato, that ``laws are imposed in the interest of rulers'', it remains true that the general conduciveness to social happiness of the habit of Order or Law-observance, is, as Hume says, too obvious to need proof; indeed it is of such paramount importance to a community, that even where particular laws are clearly injurious it is usually expedient to observe them, apart from any penalty which their breach might entail on the individual. We saw, however, that Common Sense sometimes bids us refuse obedience to bad laws, because ``we ought to obey God rather than men'' (though there seems to be no clear intuition as to the kind or degree of badness that justifies resistance); and further allows us, in special emergencies, to violate rules generally good, for ``necessity has no law'', and ``salus populi suprema lex''.
These and similar common opinions seem at least to suggest that the limits of the duty of Law-observance are to be determined by utilitarian considerations. While, again, the Utilitarian view gets rid of the difficulties in which the attempt to define intuitively the truly legitimate source of legislative authority involved us; at the same time that it justifies to some extent each of the different views current as to the intrinsic legitimacy of governments. For, on the one hand, it finds the moral basis of any established political order primarily in its effects rather than its causes; so that, generally speaking, obedience will seem due to any de facto government that is not governing very badly. On the other hand, in so far as laws originating in a particular way are likely to be (1) better, or (2) more readily observed, it is a Utilitarian duty to aim at introducing this mode of origination: and thus in a certain stage of social development it may be right that (e.g.) a `representative system' should be popularly demanded, or possibly (in extreme cases) even introduced by force: while, again, there is expediency in maintaining an ancient mode of legislation, because men readily obey such: and loyalty to a dispossessed government may be on the whole expedient, even at the cost of some temporary suffering and disorder, in order that ambitious men may not find usurpation too easy. Here, as elsewhere, Utilitarianism at once supports the different reasons commonly put forward as absolute, and also brings them theoretically to a common measure, so that in any particular case we have a principle of decision between conflicting political arguments.
As was before said, this Law-observance, in so far at least as it affects the interests of other individuals, is what we frequently mean by Justice. It seems, however, that the notion of Justice, exhaustively analysed, includes several distinct elements combined in a somewhat complex manner: we have to inquire, therefore, what latent utilities are represented by each of these elements.
Now, first, a constant part of the notion, which appears in it even when the Just is not distinguished from the Legal, is impartiality or the negation of arbitrary inequality. This impartiality, as we saw (whether exhibited in the establishment or in the administration of laws), is merely a special application of the wider maxim that it cannot be right to treat two persons differently if their cases are similar in all material circumstances. And Utilitarianism, as we saw, admits this maxim no less than other systems of Ethics. At the same time, this negative criterion is clearly inadequate for the complete determination of what is just in laws, or in conduct generally; when we have admitted this, it still remains to ask, ``What are the inequalities in laws, and in the distribution of pleasures and pains outside the sphere of law, which are not arbitrary and unreasonable? and to what general principles can they be reduced?''
Here in the first place we may explain, on utilitarian principles, why apparently arbitrary inequality in a certain part of the conduct of individuals is not regarded as injustice or even---in some cases---as in any way censurable. For freedom of action is an important source of happiness to the agents, and a socially useful stimulus to their energies: hence it is obviously expedient that a man's free choice in the distribution of wealth or kind services should not be restrained by the fear of legal penalties, or even of social disapprobation, beyond what the interests of others clearly require; and therefore, when distinctly recognised claims are satisfied, it is pro tanto expedient that the mere preferences of an individual should be treated by others as legitimate grounds for inequality in the distribution of his property or services. Nay, as we have before seen, it is within certain limits expedient that each individual should practically regard his own unreasoned impulses as reasonable grounds of action: as in the rendering of services prompted by such affections as are normally and properly spontaneous and unforced.
Passing to consider the general principles upon which `just claims' as commonly recognised appear to be based, we notice that the grounds of a number of such claims may be brought under the general head of `normal expectations'; but that the stringency of such obligations varies much in degree, according as the expectations are based upon definite engagements, or on some vague mutual understanding, or are merely such as an average man would form from past experience of the conduct of other men. In these latter cases Common Sense appeared to be somewhat perplexed as to the validity of the claims. But for the Utilitarian the difficulty has ceased to exist. He will hold any disappointment of expectations to be pro tanto an evil, but a greater evil in proportion to the previous security of the expectant individual, from the greater shock thus given to his reliance on the conduct of his fellow-men generally: and many times greater in proportion as the expectation is generally recognised as normal and reasonable, as in this case the shock extends to all who are in any way cognisant of his disappointment. The importance to mankind of being able to rely on each other's actions is so great, that in ordinary cases of absolutely definite engagements there is scarcely any advantage that can counterbalance the harm done by violating them. Still, we found that several exceptions and qualifications to the rule of Good Faith were more or less distinctly recognised by Common Sense: and most of these have a utilitarian basis, which it does not need much penetration to discern. To begin, we may notice that the superficial view of the obligation of a promise which makes it depend on the assertion of the promiser, and not, as Utilitarians hold, on the expectations produced in the promisee, cannot fairly be attributed to Common Sense: which certainly condemns a breach of promise much more strongly when others have acted in reliance on it, than when its observance did not directly concern others, so that its breach involves for them only the indirect evil of a bad precedent, as when a man breaks a pledge of total abstinence. We see, again, how the utilitarian reasons for keeping a promise are diminished by a material change of circumstances, for in that case the expectations disappointed by breaking it are at least not those which the promise originally created. It is obvious, too, that it is a disadvantage to the community that men should be able to rely on the performance of promises procured by fraud or unlawful force, so far as encouragement is thereby given to the use of fraud or force for this end. We saw, again, that when the performance would be injurious to the promisee, Common Sense is disposed to admit that its obligation is superseded; and is at least doubtful whether the promise should be kept, even when it is only the promiser who would be injured, if the harm be extreme;---both which qualifications are in harmony with Utilitarianism. And similarly for the other qualifications and exceptions: they all turn out to be as clearly utilitarian, as the general utility of keeping one's word is plain and manifest.
But further, the expediency of satisfying normal expectations, even when they are not based upon a definite contract, is undeniable; it will clearly conduce to the tranquillity of social existence, and to the settled and well-adjusted activity on which social happiness greatly depends, that such expectations should be as little as possible baulked. And here Utilitarianism relieves us of the difficulties which beset the common view of just conduct as something absolutely precise and definite. For in this vaguer region we cannot draw a sharp line between valid and invalid claims; `injustice' shades gradually off into mere `hardship'. Hence the Utilitarian view that the disappointment of natural expectations is an evil, but an evil which must sometimes be incurred for the sake of a greater good, is that to which Common Sense is practically forced, though it is difficult to reconcile it with the theoretical absoluteness of Justice in the Intuitional view of Morality.
The gain of recognising the relativity of this obligation will be still more felt, when we consider what I distinguished as Ideal Justice, and examine the general conceptions of this which we find expressed or latent in current criticisms of the existing order of Society.
We have seen that there are two competing views of an ideally just social order---or perhaps we may say two extreme types between which the looser notions of ordinary men seem to fluctuate---which I called respectively Individualistic and Socialistic. According to the former view an ideal system of Law ought to aim at Freedom, or perfect mutual non-interference of all the members of the community, as an absolute end. Now the general utilitarian reasons for leaving each rational adult free to seek happiness in his own way are obvious and striking: for, generally speaking, each is best qualified to provide for his own interests, since even when he does not know best what they are and how to attain them, he is at any rate most keenly concerned for them: and again, the consciousness of freedom and concomitant responsibility increases the average effective activity of men: and besides, the discomfort of constraint is directly an evil and pro tanto to be avoided. Still, we saw that the attempt to construct a consistent code of laws, taking Maximum Freedom (instead of Happiness) as an absolute end, must lead to startling paradoxes and insoluble puzzles: and in fact the practical interpretation of the notion `Freedom', and the limits within which its realisation has been actually sought, have always---even in the freest societies---been more or less consciously determined by considerations of expediency. So that we may fairly say that in so far as Common Sense has adopted the Individualistic ideal in politics, it has always been as subordinate to and limited by the Utilitarian first principle.
It seems, however, that what we commonly demand or long for, under the name of Ideal Justice, is not so much the realisation of Freedom, as the distribution of good and evil according to Desert: indeed it is as a means to this latter end that Freedom is often advocated; for it is said that if we protect men completely from mutual interference, each will reap the good and bad consequences of his own conduct, and so be happy or unhappy in proportion to his deserts. In particular, it has been widely held that if a free exchange of wealth and services is allowed, each individual will obtain from society, in money or other advantages, what his services are really worth. We saw, however, that the price which an individual obtains under a system of perfect free trade, for wealth or services exchanged by him, may for several reasons be not proportioned to the social utility of what he exchanges: and reflective Common Sense seems to admit this disproportion as to some extent legitimate, under the influence of utilitarian considerations correcting the unreflective utterances of moral sentiments.
To take a particular case: if a moral man were asked how far it is right to take advantage in bargaining of another's ignorance, probably his first impulse would be to condemn such a procedure altogether. But reflection, I think, would show him that such a censure would be too sweeping: that it would be contrary to Common Sense to ``blame A for having, in negotiating with a stranger B, taken advantage of his ignorance of facts known to himself, provided that A's superior knowledge had been obtained by a legitimate use of diligence and foresight, which B might have used with equal success What prevents us from censuring in this and similar cases is, I conceive, a more or less conscious apprehension of the indefinite loss to the wealth of the community that is likely to result from any effective social restrictions on the free pursuit and exercise'' of economic knowledge. And for somewhat similar reasons of general expediency, if the question be raised whether it is fair for a class of persons to gain by the unfavourable economic situation of any class with which they deal, Common Sense at least hesitates to censure such gains---at any rate when such unfavourable situation is due ``to the gradual action of general causes, for the existence of which the persons who gain are not specially responsible''. 
The general principle of `requiting good desert', so far as Common Sense really accepts it as practically applicable to the relations of men in society, is broadly in harmony with Utilitarianism; since we obviously encourage the production of general happiness by rewarding men for felicific conduct; only the Utilitarian scale of rewards will not be determined entirely by the magnitude of the services performed, but partly also by the difficulty of inducing men to perform them. But this latter element seems to be always taken into account (though perhaps unconsciously) by Common Sense: for, as we have been led to notice, we do not commonly recognise merit in right actions, if they are such as men are naturally inclined to perform rather too much than too little. Again, in cases where the Intuitional principle that ill-desert lies in wrong intention conflicts with the Utilitarian view of punishment as purely preventive, we find that in the actual administration of criminal justice, Common Sense is forced, however reluctantly, into practical agreement with Utilitarianism. Thus after a civil war it demands the execution of the most purely patriotic rebels; and after a railway accident it clamours for the severe punishment of unintentional neglects, which, except for their consequences, would have been regarded as very venial.
If, however, in any distribution of pleasures and privileges, or of pains and burdens, considerations of desert do not properly come in (i.e. if the good or evil to be distributed have no relation to any conduct on the part of the persons who are to receive either)---or if it is practically impossible to take such considerations into account---then Common Sense seems to fall back on simple equality as the principle of just apportionment. And we have seen that the Utilitarian, in the case supposed, will reasonably accept Equality as the only mode of distribution that is not arbitrary; and it may be observed that this mode of apportioning the means of happiness is likely to produce more happiness on the whole, not only because men have a disinterested aversion to unreason, but still more because they have an aversion to any kind of inferiority to others (which is much intensified when the inferiority seems unreasonable). This latter feeling is so strong that it often prevails in spite of obvious claims of desert; and it may even be sometimes expedient that it should so prevail.
For, finally, it must be observed that Utilitarianism furnishes us with a common standard to which the different elements included in the notion of Justice may be reduced. Such a standard is imperatively required: as these different elements are continually liable to conflict with each other. The issue, for example, in practical politics between Conservatives and Reformers often represents such a conflict: the question is, whether we ought to do a certain violence to expectations arising naturally out of the existing social order, with the view of bringing about a distribution of the means of happiness more in accordance with ideal justice. Here, if my analysis of the common notion of Justice be sound, the attempt to extract from it a clear decision of such an issue must necessarily fail: as the conflict is, so to say, permanently latent in the very core of Common Sense. But the Utilitarian will merely use this notion of Justice as a guide to different kinds of utilities; and in so far as these are incompatible, he will balance one set of advantages against the other, and decide according to the preponderance.