You are here

IX: Duty

1. IN the preceding chapter we introduced the notion of a well-grounded norm. A heteronomous command we call well-grounded when the act of issuing it is a practical necessity with a view to the norm-authority's ends. Autonomous norms are ipso facto well—grounded. (Ch. VIII sect. 9.)

We shall in this chapter study a special case of well-grounded norms. This is the case when the ultimate end relative to which the norm is well-grounded is the good of some being. Of this case I shall say that it imposes a duty on the norm-subject.

The term ‘duty’ is used with a multitude of meanings in ordinary language. Here it is used as a technical term. Not everything which is called a ‘duty’ is a duty in our sense. Legal duties e.g. need not be. Whether so-called moral duties are or not will depend upon the view we take of the nature of morality. But I think it is true to say that everything which is a duty in our sense i.e. is a practical necessity with a view to (promoting or respecting) the good of some being is in common speech quite naturally called a duty. Therefore our use of ‘duty’ as technical term is in good agreement with one of the uses of this word in ordinary language.

Duties in our sense can be suitably divided into certain main categories.

A first division is into self-regarding and other-regarding duties. The names and also the sense which we give to them here are familiar from traditional ethics. When a duty is self-regarding the good which the agent is supposed to serve by his dutiful action is the agent's own welfare. When a duty is other-regarding it is the welfare of some being other than the agent himself.

A second division is into autonomous and heteronomous duties.

A duty is autonomous when dutiful action is incumbent on the agent itself as a practical necessity—independently of whether the good which it serves is the agent's own or some other being's. A duty is heteronomous when dutiful action is heteronomously prescribed to the agent.

When we combine these two grounds of division we get in all four basic types of duty viz autonomous self-regarding autonomous other-regarding heteronomous self-regarding and heteronomous other-regarding duties.

There is however also a third way of dividing duties which must be noted. It is their division into that which I propose to call positive and negative duties. The first are duties to do the second duties to forbear something. The two kinds of duty thus answer to two sub-kinds of that which we have called positive and negative commands. (See Ch. VIII sect. 3.) Of the negative commands we said that they are also commonly called prohibitions.

From the dutybound agent's point of view positive duties are (mainly) duties to promote negative duties are duties to respect the good of beings. From the point of view of the norm-authority again the purpose of negative duties can be said to be (chiefly) to protect the good of some being.

Negative other-regarding duties are related to one of the many different concepts of a right. This notion of a right is a normative idea with a characteristic dual aspect. From the point of view of the right-holder a right in this sense is a freedom or permission to act in a certain way. From the point of view of the dutybound agents again the right is a prohibition to interfere with the right-holder's action should he choose to avail himself of his right.

2. An autonomous self-regarding duty according to the definitions we have given is an autonomous necessitation of the will to do something for the sake of promoting or protecting the acting agent's own good.

Autonomous self-regarding duties must not be confused with autonomous practical necessities in general. Consider once again the example of the man who runs to the station. (See Ch. VIII sect. 7.) He wants to be in time for the train. Unless he runs he will be late. Therefore he has to run. This is autonomous practical necessitation. Why is it not autonomous self-regarding duty?

Wanting to be at the station in time is usually not an ultimate end of action. A man may want this because he wants to meet someone or because he has promised to meet someone and is anxious to fulfil his promise. These could be ultimate ends. Some such ends may actually make his action other-regarding duty. They would not make it self-regarding unless the agent could say truly of himself that he wants to be at the station in time because his welfare demands this. I am not suggesting that a man could never say this truly of himself. But it would certainly be an uncommon case.

There are two main types of case when a man can be said to care for his own good and on that account to have autonomous duties towards himself.

The first case of caring for one's good gives rise to negative duties only. There are certain things which a man shuns or regards as ‘in themselves’ unwanted. He may sometimes be willing to suffer those things for the sake of some end which he wants to attain—as e.g. when he decides to have a tooth pulled out. But unless he has some such end or some other wanted consequences in view he will consider those things bad for him and try to avoid them. Now assume that he realizes that if he does p then some such in itself unwanted thing q will befall him. He finds for example that a particular kind of food which he likes in itself upsets him very violently. He could then adopt as a maxim or rule never to take food of that sort. He as it were enforces a certain prohibition on himself for the sake of protecting himself (his ‘good’) against something unwanted. Such self-protective self-prohibitions can be called autonomous self-regarding duties.

The second case of caring for one's own good occurs when a man deliberates about his ends.

But does a man ever deliberate about his ends? Can one deliberate about ends? The question was discussed with much ingenuity by Aristotle.1 On Aristotle's view deliberation is about means not about ends. So-called deliberation about ends is about intermediate ends and therefore really about means to some ulterior ends.

It is probably right to say that the word ‘to deliberate’ is most commonly used in situations when we raise such questions as ‘How shall I achieve this?’ or ‘Which way shall I choose to do this?’ When we raise such questions we are sometimes deliberating about means to given ends and sometimes about that which we have called ‘ways of doing things’. (Aristotle did not distinguish the two.) But we are not deliberating about ends.

Sometimes however a man stops to consider questions like this: ‘What consequences is my pursuit of this end going to have—beside my possibly attaining the end?’ He is perhaps working hard to get a promotion in his job. ‘Am I ruining my health? Is it good for me never to afford time for relaxation? And when I get my promotion what will then happen? I shall get a higher pay but there will also be heavier work and more responsibility and even less time than now for relaxation.’

When a man in pursuit of ends stops to reflect upon the consequences of his pursuing and/or of his attaining his ends then he can truly be said to be deliberating about ends. This he can also be said to do when without a view to existing ends he asks himself the question ‘What shall I do?’

In such situations however a man can also be said to be deliberating about his own welfare. He asks what is good and bad for him. As a consequence of such self-searching activity he may come to aim at new or revised ends.

A man who deliberates about ends may come to see that there are certain things which he ought to do and perhaps more often must not do lest he shall regret it later in life. He may come to the conclusion that he must take some physical exercise or else he will neglect his health or that he must afford some time for reading novels or listening to music or else his soul will dry up completely or if he tends to be a spendthrift that he must think of saving for his old age. In reaching such conclusions a man may be said to impose upon himself autonomous self-regarding duties. He as it were forces upon himself a certain course of action or way of living with a view to what he considers necessary for his welfare.

There cannot exist an autonomous self-regarding duty to care for one's own good (welfare) in either of the two chief ways in which such care takes place. Self-care is the foundation and source of all autonomous self-regarding duties. To say that a man imposes upon himself the duty to care for himself is to say that his care for his own good makes it necessary for him to care for his own good. This is empty talk unless it means that it imposes upon him specific self-regarding duties e.g. a duty to care that he gets enough sleep or exercise.

3. The practical syllogism which embodies the prototype of an autonomous other-regarding duty is this:

x wants to promote or respect the good of y for its own sake.

Unless x does (forbears) p he will not be promoting

or respecting the good of y.

x must do (forbear) p.

By respecting the good of another being one may understand the forbearance of any act which if done would be bad for this being. On this definition the duty to respect is necessarily a duty to forbear. By promoting the good of another being again one may understand the doing of something which will be good for this being. On this definition the duty to promote a being's good is necessarily a duty to act in a certain way.

It may however happen that a man by forbearing to act becomes responsible for damage to another man e.g. because a third man does some harm to him. Then a duty to act (to interfere) can be called a duty to protect the neighbour's good and this duty can be distinguished both from the positive duty to promote and from the negative duty to respect this good.

It may also happen that a man by forbearing to act will promote the welfare of another man e.g. because this gives the other man an opportunity which he would otherwise not have had of doing something he likes. Then a duty to forbear (to stay aside) can be called a duty to forsake and this negative duty can be distinguished from the other negative duty to respect another man's good.

There are thus in all four types of duty which fall under the case we are now discussing: the positive duties to promote and protect another man's good and the negative duties to respect and to forsake. But forsaking can also be regarded as a negative form of promoting a good and protecting as a positive form of respecting it. I shall therefore include forsaking under promoting and protecting under respecting. And I shall say that basically promoting is a positive and respecting a negative duty.

Autonomous other-regarding duties to promote the welfare of other beings have to do with the feeling or sentiment we call love in a broad sense of this word.

Other-regarding duties both autonomous and heteronomous which require us to respect the good of other beings are felt to have a peculiar connexion with morality to be the moral duties par excellence. We shall in the last chapter examine this opinion more in detail.

We distinguished in a previous chapter between self-and other-regarding virtues. To try to acquire the other-regarding virtues: consideration helpfulness honesty etc. can be part of the autonomous other-regarding duties of a man. To try to acquire and observe them is certainly part of the heteronomous other-regarding duties which we wish to implant in others. Of the other-regarding virtues most it would seem have to do with respecting our neighbour's good; but some such as helpfulness are obviously relevant to the promotion of this good.

When speaking of self-and other-regarding virtues (see Ch. VII sect. 10) we noted that practising the latter presupposes an interest in the welfare of other people. This is the interest to which the first premiss of our practical syllogism above refers. Whether a man takes such an interest in another man or in other men generally is contingent. For this reason there can exist no such thing as an autonomous duty to love our neighbour or to want to respect his good for its own sake. Love of our neighbour is the foundation and source of any autonomous other-regarding duty we may have i.e. impose upon ourselves to promote our neighbour's good. Similarly the will to respect his good for its own sake is the foundation and source of every autonomous other-regarding duty to respect it in this or that particular regard. Interest in another man's good for its own sake cannot be autonomous other-regarding duty since it is presupposed in every such duty.

But is such interest even possible albeit not duty? Here we are touching upon a major problem of ethics viz. the problem of egoism and altruism.

4. By interest in or regard for somebody's good I shall here understand a desire to promote or respect this good.

That promoting his neighbour's welfare can be an intermediate end of a man's action presents no problem. This case simply amounts to that regard for his neighbour's welfare is a means to some ulterior end of the agent's. If the means is necessary regard for his neighbour is forced upon the agent by autonomous practical necessity.

There are innumerable ways in which regard for the good of others can become a means to self-interested ends. The master wants his servant to work effectively for him. This the servant will not do unless the master cares for his welfare—not to speak of the necessity of not maltreating the servant. Ergo will the master take heed to respect and promote the welfare of his servant to the extent he thinks necessary. This is not what I here call autonomous other-regarding duty since the ultimate end which necessitates action is not an interest in the welfare of another being.

By the thesis of Psychological Egoism one could understand the idea that promoting and respecting his neighbour's good is never anything but (at most) an intermediate end of a man's action. Regard for the good of others for its own sake we could call pursuit of an altruistic end. Egoism as here defined is thus a doctrine which denies the existence of altruism i.e. of altruistic ends of action.

The nature of this negative thesis must not be misunderstood. Egoism as here defined does of course not deny that promoting and respecting another man's good can be an end of action. It can even have the appearance of being an ultimate end. But egoism maintains that if such an end is not overtly intermediate it can become unmasked as an intermediate end.

The thesis of egoism however admits of several interpretations—which supporters of it have not always kept well apart. The denial of altruism can be understood as the denial of a logical possibility or as the denial that something is a fact. It is egoism as a denial of the logical possibility of altruistic ends of action which here interests us first of all.

It is easy to see I think what is the nerve of the idea that action cannot be genuinely altruistic. It is a logical fact that any end of action whatsoever is some agent's end i.e. is something which the agent whose ends we are considering wants to attain. Therefore if promoting my neighbour's welfare is an end of my action it is yet an end of my action something I want my interest for me a good. There is a sense in which aims in acting are helplessly self-centred.

The thesis of egoism which we are now discussing rests on a misunderstanding of the significance of the logical facts which we have just mentioned about ends of action wanting and the good of a being. Egoism misconstrues the necessary connexion which there exists between my neighbour's welfare as an end of my action and my welfare as an impossibility of pursuing the first except as a means to safeguarding or promoting the second. This is the very same mistake as the one which psychological eudaimonism commits. (Cf. Ch. V sect. 2 the discussion of the example on p. 90.) It is related to the self-refuting mistake which psychological hedonism commits when it misinterprets the necessary connexion between satisfaction of desire and pleasure as meaning that pleasure is necessarily the ultimate object of every desire.

Psychological hedonism egoism and eudaimonism are closely related philosophical views of human nature. They are all false. They all commit very much the same mistake. Their mistake consists in a misinterpretation of an existing necessary connexion. These necessities ‘behind’ the falsehoods give to the three doctrines the strong appearance of truth which they possess. A refutation of hedonism which denies the necessary connexion between desire and pleasure and a refutation of egoism or of eudaimonism which denies the necessary self-centredness of all ends of action is therefore doomed to fail of its object. The aim of the refutation is not to deny these connexions but to put right their significance.

When the logical mistake of egoism has been corrected the logical possibility of altruistic (ends of) action has become established.

The admission of the logical possibility of altruism however is fully compatible with the view that as a matter of fact regard for our neighbour is never an ultimate moving force of human conduct. Apparent altruism on this view is always egoism in disguise.

We need not here discuss this doctrine in detail. It contains a good portion of truth. Its claim to the whole truth is founded I think partly on an exaggerated pessimism about human nature and partly on the logical mistake which we just mentioned relating to the self-centredness of ends of action. In the gloom of pessimism it becomes tempting to give to one's insight into the vileness of human nature an absolute character which can be given to it only at the expense of committing a logical mistake.

Beside the logical question whether a man can take an ultimate interest in the good of his neighbour and the psychological question whether a man ever is interested in his neighbour's good except as a means to some end there is also the genetic question whether every interest in another man's good as an end was not in origin an interest in this good as a means.

Those who take the view that altruism must have a genetic foundation in self-interest are somehow astonished at the fact that there should exist such a thing as pure unselfishness i.e. action which springs from affection friendship love sympathy or respect and which is completely untinged by calculations about its utilitarian value for the agent. That man acts self-interestedly seems not to require any explanation. But that he acts unselfishly may strike us as something of a ‘mystery’.

If unselfishness a ‘mystery’ once we have seen clearly through the logical mistake which egoism commits? I do not know what is the right answer to the question. The answer of course partly depends upon what we mean by a ‘mystery’ here. But I tend to think that the appearance of ‘mystery’ or of something standing in need of ‘explanation’ which pure unselfish action can be said to exhibit is really nothing but a conceptual confusion nourished by the same mistake as the one underlying hedonism and eudaimonism. Therefore I am also inclined to think that action inspired by affection friendship and love is just as typical of human nature as action in a spirit of self-interest and that the view of man as essentially a self-interested creature is a logical misconception. But I may be mistaken.

5. Behind heteronomous self-regarding duties there is a practical syllogism of the following type:

x wants the good of y to become promoted for its own sake.

Unless y does p his good will not be promoted.

x must make y do p.

The distinction between promoting and respecting and also the distinction between doing and forbearing is not it seems of much importance to the case which we are now considering. I shall therefore here use the verb ‘promote’ to cover ‘promote or respect’ and the verb ‘do’ to cover ‘do or forbear’.

The conclusion of the above syllogism is an autonomous practical necessity incumbent upon x. x ought to make y do p. There are many ways in which x can try to achieve this. One way is by commanding y. This is the only way which interests us here. When it is resorted to then x imposes upon y a duty to act in a manner which x considers good for y. The duty is heteronomous. y or the norm-subject is not moved to action by his own wants and insights into necessities. It is x or the norm-authority who is thus moved.

Thus when there is a heteronomous self-regarding duty the norm-subject's good is the norm-authority's end. The authority imposes upon the subject a duty for the sake of (in the name of) the subject's welfare. But the subject if he obeys does not necessarily do the ordered thing for the sake of his good. He may do it simply and solely for the sake of escaping punishment. Or he may do it because he wants to please the authority or out of respect for the authority.

6. When does commanding actually take place ‘in the name of the good of’ the commanded? There are some typical cases. One is when parents order their children to do certain things because it is good for the child or to forbear something because it would be bad for the child to do it. Also in the relations of teachers to pupils orders of a similar nature are sometimes given. Finally people who are feebleminded or for some other reason incapable of looking after themselves may have to be commanded by others who have a better understanding of their welfare.

It will be noted that the agent who issues commands to others in the name of their good usually enjoys some recognized position of ‘authority’ over those whom he commands—such as parent or teacher or guardian. It will also be noted that two of the three typical cases which we mentioned have to do with that which we call education.

I shall call—partly for want of a better name—commands which are issued in the name of the good of the commanded moral commands. Of the educational purposes which such commanding may serve I shall speak as moral education. And the authority who gives such orders I shall call a moral authority.

This use of the adjective ‘moral’ incidentally is not unnatural or at great variance with ordinary usage—though maybe not very familiar in moral philosophy. It has no immediate connexion with so-called moral goodness nor with so-called moral duty. I am not suggesting that the duties imposed by what I call moral commands are those which we ordinarily call moral duties. The activity however which in common speech is called moralizing largely consists in advising and recommending to other people things in the name of their good and also in drawing their attention to various ways in which people may have neglected their own welfare. Moralizing is not co-extensive with that which I here call moral commanding. But that which I here call moral commanding falls under that which is ordinarily called moralizing.

We shall distinguish between the reasons and the justification for commanding others in the name of their good.

The immediate reason for such commanding is that the norm-authority takes an interest in the welfare of the norm-subject. The immediate reason need not however be an ultimate reason. Sometimes a man takes an interest in the good of another man for the sake of some ulterior end of his own. This is the case of the master who cares for the good of his servants because he expects a better return of services. Sometimes a man is under a legal obligation to care for the good of others in a manner which involves commanding them. Parents and guardians have such obligations. It is an important though not the sole aspect of parental love that this love should be the ultimate reason why parents oblige their children to acts and forbearances in the name of the children's good.

Only when the ultimate reason for commanding others is an interest in their welfare shall we say that a heteronomous self-regarding duty is being imposed.

Commanding others in the name of their good can be said to entail a claim on the part of the norm-authority to a better insight into the requirements of the welfare of the norm-subject than the norm-subject can claim for himself. The justification of this claim is part of the justification of heteronomous self-regarding duties.

It is a fact that some men have a much less developed conception than other men of that which is good or bad for themselves. This can be due to various reasons e.g. to lack of experience of that which may befall a man as a consequence of his actions. Small infants can be said to have as yet no conception at all of that which is good or bad for them—Children who already have some views in the matter may nevertheless be completely mistaken in their views e.g. strongly want to do things which they will regret thinking that they will not. It is in such situations when a being's conception of his own welfare either is lacking or undeveloped or distorted that men are called upon to exercise moral authority over others.

Part of the justification of heteronomous other-regarding duties is thus the superior wisdom of some men as to that which is good or bad for other men. The moral authority of parents guardians and teachers can normally be expected to possess this justification for its exercise.

If men necessarily pursued their real good any claim to superior wisdom on the part of others would necessarily lack justification—and therewith also any attempt to impose heteronomous self-regarding duties. As a matter of fact however men often pursue things which ‘on second thoughts’ they consider unwanted. The test which shows that something was a constituent of a man's apparent as opposed to his real good is that he regrets it. (See Ch. V sect. 14.) To know the requirements of another man's welfare better than he does it himself is essentially the superior capacity of seeing what a man will later regret having done or neglected. The man who exercises moral authority over another is justified in doing so only to the extent that he can say truthfully to the other man: “There will be a day when you will have reason to be grateful for that which I did to guide you.” The important point is not whether the other man acknowledges his gratitude. Some men are by nature little inclined to be grateful even when there is a reason. The important point is that there should be reason for thinking that had the other man not obeyed the commands he would have been worse off in the long run or for thinking that had he obeyed he would be better off than he is now. But what is ‘better off’ and ‘worse off’ is in the last resort for the subject himself to decide.

When the claim to superior knowledge of good and evil lacks justification the exercise of moral authority of one man over another deteriorates into something which may be called moral tyranny. The subject is then forced to do things ‘in the name of his good’ when actually he knows better himself what is good or bad for him. To know whether the claim to superior knowledge on the part of the norm-authority is justified or not can be most difficult. The struggle for moral freedom i.e. for the right to act in accordance with that which one considers good or bad for oneself against alleged moral tyranny can greatly upset the relations between parents and children in the period before the latter come of age.

It is an important aspect of that which we call a child's ‘coming of age’ that the child reaches that which we shall here call moral maturity. An individual can be said to have reached moral maturity when there is no obvious justification for a claim which others may put forward to superior insight into the requirements of that individual's good.

With the reaching of moral maturity the moral education of men—in the form of heteronomous self-regarding duties being imposed on them—comes to an end. It is the aim of moral education not only to promote and protect the good of others during their moral immaturity but also to facilitate the process of maturation by teaching them to care for their own welfare. But here again it is important to remember that the criterion of the present superior wisdom of the teacher is set by the future verdict of the pupil on the fruits of his teaching.

7. We now proceed to heteronomous other-regarding duties.

A man x commands another man y to do something. The command let us assume is well-grounded. This means that x must issue it in order to secure some end of his. If this end is the good of y and if it is an ultimate end then the command imposes a heteronomous self-regarding duty on y. If the end is the good of some being other than y and is an ultimate end then the command imposes a heteronomous other-regarding duty on y in relation to that other being.

The case when one man commands a second man in the name of the welfare of a third man is a thoroughly possible and by no means unrealistic case. Nevertheless it seems to be a case of rather subordinate interest. We shall therefore here disregard it.

With this exclusion the remaining cases of heteronomous other-regarding duties are those in which one man for the sake of his own welfare commands another man. Such duties must be distinguished from heteronomous commands for self-interested ends generally. We do not wish to say that we are imposing a heteronomous other-regarding duty on a man if we order him to open a window because we want to have the room ventilated. But we may wish to say that an order not to disturb us when we are working aims at imposing an other-regarding duty.

As observed in sect. 2 one way in which a man can care for his own welfare is by taking various self-protective steps in order to escape things which he shuns as being unwanted either in themselves or due to their consequences. It would seem that it is this way of caring for one's own good that is predominantly relevant to heteronomous other-regarding duties. Such duties when imposed by one man on another in the interest of the first are mainly negative duties i.e. prohibitions to do things which the first man considers harmful for himself. From the norm-authority's point of view heteronomous other-regarding duties are chiefly self-protective prohibitions.

8. We have so far been talking as though as a matter of course any man could if he wanted to command any other man. The recipient of an order may disobey the order nevertheless he has been commanded.

As a matter of fact men do not often command other men. When this happens it usually takes place in circumstances under which commanding has become somehow ‘institutionalized’. Parents command their children but not the children of other parents. Teachers sometimes command their pupils; for pupils to command their teachers is out of place. Officers of superior rank command officers of inferior rank; commoners do not often command one another. When I ask my friend to shut the door I do not command him; not to speak of the case when I ask a stranger for a service. When I order a book through my bookseller this is not ordering in the sense of commanding.

Why is it that we do not command our friends or strangers? Is it because we are too polite thus to intrude upon them? Can we command our friends? The question is worth asking.

The fact that commanding usually takes place under some ‘institutionalized’ circumstances may suggest to us the idea that in order to be able to command one must possess something which could be called a right to command enjoy some recognized position of authority. Parents have a right to command their children children no right to command their parents. Has the highwayman who commands ‘hands up’ a right to command this? Certainly not. Is he then not commanding ‘Hands up’? He certainly is.

We have said before (Ch. VIII sect. 3) that to a command is essentially tied a threat of punishment in case of disobedience. This threat must not be empty or a mock threat. It must be an effective threat. That it is effective does not mean that it necessarily secures obedience. Nor does it mean that disobedience is without exception punished. I shall not here discuss in detail what constitutes the efficacy of a threat of punishment.

It thus follows that the possibility of commanding is essentially tied to the possibility of effectively threatening the recipient of the order with punishment in case of disobedience. This second possibility again depends upon that which I shall call the relative strength of the agents concerned. Effective threatening is on the whole possible only when the individual who threatens is stronger than the individual who is being threatened. Thus the possibility of commanding is founded upon the superior strength of the commander over the commanded.

This concept of strength stands in need of elucidation. It is not the same as that which is ordinarily understood by ‘physical strength’ though it is related to it. A big strong man may not be able to command a small weak one because the second can run away and put himself out of reach of the punishing arm of the first. But then it must be remembered that running too is part of a man's physical powers.

Our strong man can of course continue to shout out commands to the small man and even accompany his orders by the most fearful threats. Is he then not ‘issuing commands’? If he is not convinced that his words will not impress the small man he can rightly be said to be trying to command him. If he knows that his shouting and threatening is all in vain we could say that he is not commanding but only giving vent to his wishes and his anger. We could also if we wanted distinguish between commanding which would consist merely in the use of words combined with a certain intention and effectively commanding which would combine intention and the use of words with an effective threat of punishment. Then we could say that it is only effective commanding which constitutes normative relationships of the kind we are here studying between men.

The notion of strength which is essential to the possibility and existence of normative relationships among men I shall define as capacity of affecting through one's action the good of other beings favourably or adversely. By saying that two agents are of roughly equal strength we here mean that they have roughly the same capacities (powers) of influencing the welfare of one another. By calling one stronger than another we mean that the first can do more for promoting but also more for injuring the good of the second than the second can do with regard to the good of the first.

It is obvious that capacity of favouring and capacity of injuring someone's welfare need not be proportionate. The fact that one man can do more good to another than this other man can do to him does not exclude the possibility that the second man can do more harm to the first than the first can do to the second.

To the capacity of commanding the capacity of visiting another man with evil is of more importance than the capacity of doing good to him. This is so because commanding is essentially connected with punishing disobedience but not with rewarding obedience. The superior strength necessary for commanding is therefore essentially a superior power of causing suffering to others. That such should be the logical foundation of heteronomous law and heteronomous duties may seem brutish but must I think be accepted as fact.

One man may accidentally enjoy superior power or strength in the sense defined in relation to another man. Such accidental superiority can then be used for commanding. Blackmail is a species of commanding which is based upon accidental superiority of strength.

Consider two men ‘by themselves’ i.e. in abstraction from such relationships which they may have to other men and which may be relevant to their capacity of doing good or harm to one another. For example: the one man is king and can command his bodyguard to seize the other man if he does not obey the king's orders. Then abstract from this and think of the king without his bodyguard. Abstract also from such accidental superiority which one man may have over another. For example: the one man has a rifle. Abstract from this or endow in thought the other man with a rifle too. The balance of strength which remains after these abstractions I shall call the natural relative strength of the two men.

Such logical fictions as this about the powers of men in a state of nature may strike one as old-fashioned and unrealistic. They were much in favour with writers of the 17th and the 18th centuries. They may be grossly unrealistic as hypotheses concerning the conditions which existed before the conventions customs and norms of society had ‘perverted’ the ‘natural’ ways of life. As abstractions for the purposes of studying concepts these fictions may be both legitimate and useful. It is to be regretted that an exaggerated respect for the ‘naturalistic’ study of man by anthropologists historians and psychologists has made philosophers generally abandon these fictions of which the supreme product is the Contract Theory of the State.

It is a fact of fundamental importance alike to moral legal and political philosophy that men are by nature roughly equally strong i.e. endowed with roughly the same natural powers of affecting the welfare of one another favourably and adversely. I shall express this fact by saying that men are by nature approximate equals.

There are of course noteworthy natural inequalities too. Some of these inequalities are founded in physical others in intellectual superiority some in both.

There is first of all the physical and intellectual superiority of adult people over children before the latters’ coming of age. On it the normative relationship between parents and children is ultimately based. It should be noted that the natural limits to the normative authority of a father or mother over children other than their own are not set by the children of other parents but by the parents of other children.

There is further a certain inequality at least in physical strength between the sexes. Whether it is relevant to normative relationships is arguable. It certainly has been considered relevant in some quarters.

Here must also be noted the fact that physical strength up to a point declines with increasing age. This decline however is up to a point compensated for by an increase in the mental power which we call ‘wisdom’—maybe not in matters purely intellectual but certainly in matters pertaining to the good of man.

Of the normative relationships between parents and children it is characteristic not only that it exists ‘naturally’ i.e. on a basis of natural superiority in strength but also that it dissolves ‘naturally’ i.e. as a consequence of the children's growing up to become the approximate equals of their parents. It is an interesting and deeply significant fact about man that children become their parents’ equals in strength roughly at the time when they reach that which I called moral maturity i.e. become capable of caring for their own welfare. The basis of parental normative authority thus vanishes roughly at the time when its exercise loses its justification.

At least as important as the fact that men are approximate equals by nature is the fact that men can add to their strength by joining forces with other men. The ability to co-operate for ends is not a uniquely human ability. But the role which this ability plays in the creation of normative relationships among men has no counterpart elsewhere in the animal kingdom. For it is on this possibility of joining forces that much of social life on the human level depends and above all the great fabric of commanding power we call the state.

The significance for normative relationships of man's capacity to join forces with other men is twofold. Firstly it can overrule all natural (and accidental) inequalities which may exist among individual men. If there are two men x and y on a desert island and x is the stronger there may also develop between them a relationship which we could call that of master to servant. x commands and y on the whole obeys. If there are three men x and y and z and x is strongest x may succeed in making himself master of both y and z. His superior strength may partly consist in his skill to kindle unfriendly relations between y and z and thus prevent them from joining forces against him. But if y and z join forces it is most likely that they shall be able to withstand any attempt on the part of x to tyrannize over them.

Here we find the second respect in which man's ability to cooperate is important. It can create ‘artificial’ inequalities on a basis of ‘natural’ equalities. It is on such created inequalities that institutionalized normative power of some men over other men mainly rests. The ‘ruler’ in the widest sense of the word is not necessarily stronger than each of his subjects. He is almost certainly weaker than even two of them jointly. Yet he can command each of his subjects because should one of them be recalcitrant and refuse to obey orders there will be others who join hands with the ruler to punish the recalcitrant.

Thanks to the fact that men are approximate equals by nature the possibilities of the human individual to command and therewith also to impose heteronomous duties on other individuals are narrowly restricted. But thanks to the power of co-operation heteronomous normative relationships yet play an all-pervading role in the life of human communities.

Inequalities of strength based on men's capacity to co-operate are the foundation of legal duties and the reign of law. These inequalities however may also serve the establishment of moral duties and the reign of justice. How this happens I shall try to show in the next chapter.

  • 1. See EN, Bk. III, Ch. 3.