1. IN this chapter I shall discuss three typical attributions of goodness: the good act the good intention or will and—quite briefly—the good man.

It would seem that ‘good’ in the phrase ‘a good act’ unlike ‘good’ in such phrases as ‘a good knife’ or ‘a good general’ does not signify a goodness of its kind. (Cf. Ch. II sect. 1.) There is no excellence which is typical of acts as acts as there is an excellence of knives as knives and of generals as generals. But acts can participate in several forms of goodness.

There is first to be noted a connexion between acts and technical goodness. This last as we know is primarily an attribute of agents who are good at some activity. It is in the last resort in the skilful performance of some acts that technical goodness reveals itself. Acts are sometimes judged good or bad on the basis of the perfection of their performance. ‘Well done’ we say of the good performance and a well done so-and-so we sometimes call a good so-and-so e.g. when speaking of a good ski-jump or a good race.

The goodness which we attribute to an act on the basis of the excellence of its performance I shall call the technical goodness of the act. It is an attribute of an act as a member of a kind of act. It is typically a goodness of its kind. The good ski-jump is as ski-jump good. If there existed standards of excellence of performance for all kinds of acts then one could in a secondary sense call the goodness of an act's performance the goodness of the act qua act as a good golfer might also be called a good sportsman.

It should be noted that technical goodness as an attribute of acts is necessarily an attribute of act-individuals i.e. individual performances of acts of some kind and never an attribute of act-categories or kinds of act.

2. One of the most common and familiar uses of ‘good’ is its use in the phrase ‘a good (bad better less good) way of doing something’. Here ‘doing something’ stands for an arbitrary human act. An act is the bringing about or production at will of a change in the world e.g. the unlocking of a door. The change brought about or effected we call the result of the act e.g. the fact that a certain door which was locked is now open. The ‘way of doing’ again is some act or activity which leads up to’ the result of an act e.g. the turning of a key and pulling of a handle which opens the door. The tie between ‘way of doing’ and ‘thing done’ is an intrinsic connexion between a kind of act or activity and some generic state of affairs. Certain turnings of a key is a way of unlocking a door only provided it results in the door's being unlocked.

Value-judgments which are passed on ways of doing things are usually comparative judgments. That one way of doing something is better than another way of doing the same thing can mean that it is easier or quicker or less expensive or more pleasant or more tidy or is not connected with certain unwanted side-effects etc. What ‘better’ means in the individual case depends upon the aim of the doer whether he wants to do the thing as cheaply as possible or as quickly or as tidily or what not. This relativity of the goodness of a way of doing to aims of the agent does not however mean that ‘good’ here is an attribute of the individual performance. The very phrase ‘way of doing’ indicates generality.

It should be observed that also a bad way of doing a thing is a way of doing it just as a poor or a bad knife is still a knife. This means that there is a sense of ‘good’ in which even a bad way of doing something is good viz. ‘good’ as opposed to ‘no good’ for the purpose of effecting a certain result. Ways of doing which are no good in a certain acting situation are (usually) not ways of doing the wanted thing at all—and not bad ways of doing something. In all these respects the use of ‘good’ and ‘bad’ and ‘better’ as attributes of ways of doing resembles the use of ‘good’ and ‘poor’ (sometimes also ‘bad’) and ‘better’ for rating instruments.

3. The attribution of instrumental goodness to ways of doing things must not be confused with the attribution of that which we have here called utilitarian goodness i.e. roughly speaking usefulness to acts as means to certain ends.

Means-end relationships which are relevant to judgments of utilitarian goodness are causal or extrinsic and not logical or intrinsic relationships. (Cf. above Ch. III sect. 1 and below Ch. VIII sect. 5.) An act is good or useful for a certain purpose or with a view to a certain end if the doing of this act promotes or favours this purpose or the attainment of this end. It is bad or harmful if it hinders or counteracts the attainment of the end. For example: to disobey the doctor's orders may be a hindrance to quick recovery from an illness. It is a bad thing to do with a view to one's recovery. But we do not except ironically call it ‘a bad way of recovering’.

‘Good ways of doing things’ and ‘bad ways of doing things’ I would say are opposed to one another as contradictories rather than as contraries. But utilitarian goodness and badness as attributes of acts denote contraries and not contradictories. The good act is useful by favouring the bad act detrimental or harmful by counteracting the attainment of the end. These clearly are contrarily related alternatives. Between them fall the acts which neither favour nor counteract the attainment of the end.

It is useful to distinguish between the result and the consequences of an act. The result of an act is that state of affairs which must obtain if we are to say truly that the act has been done. For example: the result of the act of opening a window is that a certain window is open (at least for a short time). The consequences of an act are states of affairs which by virtue of causal necessity come about when the act has been done. (Ch. V sect. 8; see also the discussion in Ch. VIII sect. 5.) For example: A consequence of the act of opening a window may be that the temperature in the room goes down. The relation between an act and its result is intrinsic; the relation between an act and its consequences again is extrinsic.

It follows from what has been said that whether an act is good or bad in a utilitarian sense depends upon its consequences. Which will be the consequences of an act again largely depends upon the circumstances under which the act was done. The circumstances may vary from one individual performance to another of an act of a certain category or kind. For this reason attributions of utilitarian goodness to acts will often be restricted to act—individuals. But they are not necessarily thus restricted. Sometimes the consequences of individual performances of an act are nearly always the same independently of variations in circumstances. Then utilitarian goodness may become attributed to the act-category. For example: medicines and drugs have nearly uniform effects on the human body. Hence such general judgments become possible as say that taking Vitamin C is good (useful) for preventing colds.

Means are not clearly distinguished in language from ways of doing things. Sometimes it would seem the phrase ‘a good (bad) means’ means exactly the same as ‘a good (bad) way of doing’. But sometimes ‘good (bad) means’ refers to some act which is causally connected with some end. In the first case ‘good’ in ‘good means’ connoted instrumental in the second case utilitarian goodness.

It is of some interest to note that whereas technical goodness can be attributed only to individual acts viz. as performances and instrumental goodness only to kinds of act or activity viz. as ways of doing utilitarian goodness may become attributed either to an act-individual or sometimes to an act-category.

The utilitarian goodness of an individual act or of an act-category depends solely upon a causal relation to some end of action. It is completely independent of any goodness which the end may possess in addition to being ‘a good’ by virtue of being an end. There can be good means to bad ends and bad means to good ends just as there can be good ways of doing bad things and bad ways of doing good things.

4. Are acts and activities rated as good or bad according to whether they are wanted (welcomed) or unwanted (shunned) in themselves?

Action can be rated as wanted or unwanted in itself by the agent who performs it but also by various subjects who ‘suffer’ it i.e. whose lives are affected by it. The ratings by different subjects of one and the same individual act or of one and the same category of act can of course be different.

That an act or activity is in itself wanted by the agent means that it is something which he its causal requirements and consequences apart would want to do for its own sake. If an act is in itself wanted then its result is a thing wanted in itself. If again an activity is in itself wanted then practising it is a thing wanted in itself.

Practising an activity for its own sake is frequently called by the agent a nice thing or a fine thing or a lovely thing to do. Sometimes it is also called by him a good thing to do. But in calling it good he would usually I think be implying that he considers this thing good for him i.e. beneficial.

That which an agent does is not very often in itself i.e. its consequences apart a wanted thing for another subject. But it is not infrequently in itself unwanted by others. They find it say annoying to hear or disagreeable to watch. Action on the part of other subjects which we consider in itself unwanted is commonly said to be a nuisance (to us). But we would hesitate to call it bad unless we consider it positively harmful obnoxious (for somebody).

By the intrinsic value of acts and activities we may understand their character of being wanted or unwanted or indifferent in themselves. Accepting this terminology we seem entitled to say that judging action good or bad on the basis of its intrinsic value is neither a common nor a very important kind of valuation.

Common however is the rating of acts and activities as good or bad according to whether they are (thought) beneficial or harmful. The agent of a beneficial act is said to do good to the subject or subjects for whom the act is beneficial. Similarly the agent of a harmful act is said to do barm to the subject(s) for whom the act is harmful. Also the acts themselves are sometimes said to do good and harm respectively to the subjects. More commonly however they are said to be good for and bad for those whom they affect.

One and the same (individual) act can be beneficial for one but harmful for another subject. Such acts are sometimes said to be both good and bad. There is no contradiction in this as long as we remember the relativity of the beneficial and the harmful to (‘suffering’) subjects.

An agent can be mistaken in thinking that an act of his is beneficial (harmful) for some other subject. This he can be because he is ignorant of the (causal prerequisites and) consequences of his act. But he can also be mistaken because he is ignorant of the valuation of his act by the other subject. Perhaps he did something which he was sure his neighbour would welcome. Perhaps he did this even for the sake of promoting his neighbour's good. Yet if his neighbour sincerely resents what was done to him including its consequences the act was not beneficial.

Similarly a subject can be mistaken in judging of something which has been done ‘to him’ that it is good (bad) for him. He can be mistaken because he is ignorant of the consequences but also because he fails to anticipate correctly his own valuation of the causal whole of which the act in question is the ‘nucleus’.

Judgments of the goodness or badness of human acts according to how they affect the good of various beings thus share in the precariousness and uncertainties of judgments of things beneficial and harmful in general.

5. Human acts are perhaps the most important category of things which are judged good or bad ‘in a moral sense’ or ‘from a moral point of view’.

Is there a special moral sense of the word ‘good’? Is moral goodness a special form of the good—on a par with hedonic technical utilitarian etc. goodness?

Moral philosophers often discuss the good and the bad (evil) as though the answer to the above questions were affirmative. Whether they are prepared explicitly to defend such a view is usually not clear from the very scanty treatment which those philosophers give to the other forms of goodness beside the moral. There is a tendency to dismiss the other forms or some of them as ethically irrelevant.

My own view of the matter is roughly as follows: Moral goodness is not a form of the good on a level with the other forms which we have distinguished. If it be called a form of goodness at all it is this in a secondary sense. By this I mean that an account of the conceptual nature of moral goodness has to be given in terms of some other form of the good. (Cf. Ch. I sect. 8.)

I shall here attempt to give such an account. The form of goodness in the terms of which I propose to explicate the notion of the morally good is the sub-form of utilitarian goodness which we have called the beneficial. To put my main idea very crudely: Whether an act is morally good or bad depends upon its character of being beneficial or harmful i.e. depends upon the way in which it affects the good of various beings.

There is a prima facie objection to an account of moral goodness in terms of the beneficial. It can be framed as follows:

The beneficial and the harmful are relative to subjects. If an act is called good on the ground that it is beneficial the judgment is incomplete unless we are told for whom it is good (beneficial). Similarly if an act is called bad on the ground that it does some harm the statement is elliptic unless it is added for whom the act is bad. ‘Good’ when it means ‘beneficial’ is always ‘good for somebody’ and ‘bad’ when it means ‘harmful’ is always ‘bad for somebody’.

The morally good and bad is not in this sense relative to subjects. Phrases such as ‘morally good for me’ or ‘morally bad for him’ must be dismissed as nonsensical. The fact that an act does harm to somebody may be relevantly connected with the moral badness of the act. But if this act is morally bad then it is bad simpliciter—and not for some subject as opposed to others. And similarly for moral goodness.

There is thus one sense in which moral goodness is ‘absolute’ and ‘objective’ and in which the beneficial is ‘relative’ and ‘subjective’. This marks an important logical distinction between the morally good and bad on the one hand and the beneficial and the harmful on the other hand. Does not this difference between the two ‘forms’ of goodness doom to failure any attempt to define moral goodness in terms of the beneficial? I hope to be able to show that this is not so.

6. One and the same individual act we said (p. 118) can affect the good of different beings differently—be beneficial for some and harmful for other beings. The question may be raised whether taking into consideration all such effects one could form a resultant judgment of the value of the act. If the answer is affirmative the further question could be raised whether this resultant value of the act its ‘overall character’ of beneficial or harmful could be identified with its moral worth.

The question whether a value-resultant can be formed on the basis of the character of an act of being beneficial for some and harmful for other beings is related to the following more general problem: Can the welfare of one being be ‘balanced’ against the welfare of another? Does it for example make sense to say that the good which an act did to some person was greater than the harm it did to another? Or to say that the total amount of moderate good which an act does to several beings ‘outweighs’ the great amount of harm which this same act does to one or a few beings?

It seems to me obvious that if a value-resultant of the ways in which acts affect the welfare of various beings can be formed at all this can only be done by judging the beneficial or harmful nature of the act from the point of view of the welfare of a collectivity of which the various beings concerned are the individual members. Such judging of acts is as a matter of fact often attempted. It is attempted e.g. when we argue that it is better that one man's interests are sacrificed than that all the members of a community shall suffer. Or when we say that it is better that one man is thrown overboard than that the whole crew shall perish. We are then as it were saying: this is how the community would prefer to have it. We compare the community to a man who ponders whether he should undergo a painful operation or do something he dislikes for the sake of his subsequent health or well-being.

Are such arguments logically legitimate? The answer is tied to the problem whether the notion of the good (welfare) of a collectivity or community of men is logically legitimate. I am not suggesting that the answers must be negative. But I am sure that the conditions of estimations of value from the point of view of the good of a community are extremely complicated and also that the appeal to the welfare of a collectivity over and against the welfare of some of its members is often misused in practice. We are here touching upon a major problem in the ethics of politics. I shall not discuss this problem further in the present work. I leave it open.

In the subsequent discussion I shall disregard the possibility of forming a judgment as to the ‘overall’ beneficial or harmful character of an act. I shall discuss the notion of moral goodness from what could also be called the point of view of the human individual as distinct from the point of view of the human community. On the logical complications connected with the second point of view I shall not here touch nor on the question whether the two points of view can be brought into harmony with one another.

7. We next put forward the following suggestions of how the moral value of a human act may be considered a ‘function’ of the way in which this act affects the good of various beings favourably or adversely:

an act is morally good if and only if it does good to at least one being and does not do bad (harm) to any being; and

an act is morally bad if and only if it does bad (harm) to at least one being.

The suggestions are open to a number of objections.

First it may be objected that an act could not be called morally good on the ground that it is beneficial for the agent himself—even if it does not hurt anybody else. Similarly it is at least doubtful whether harming oneself could be considered morally bad.

If one accepts one or both of these objections as valid one takes the view that moral action is essentially ‘social’. On this view that part of a man's action which affects solely his own good is morally irrelevant.

There is a certain inclination it seems to me to say that harming oneself is morally bad though doing good to oneself is not morally good. It may be suggested however that the foundation of this inclination is the fact that by harming himself a person can hardly fail to become a nuisance to or a burden on his fellow humans. According to this suggestion the moral badness of doing bad to oneself consists in the bad which the agent (indirectly) causes to others.

We shall not here take a stand on the issue whether action which solely affects the acting agent's own good is morally relevant or not.

Another objection to the suggested definition of moral goodness runs as follows: An act which does good to some and bad to no being need not have been done for the sake of doing good. The agent may not even be aware of the beneficial nature of what he did. Would it not in either case be absurd to label this act morally good?

A similar objection may be raised against our proposed definition of moral badness. If the agent is not aware of the harmful nature of what he did is his act then morally bad?

These objections mean that our proposed definition of moral goodness was too ‘lenient’ and our proposed definition of moral badness was too ‘severe’. The first was too lenient because it was compatible with the possibility that action can be morally good though no good is intended or even foreseen by the agent. The second again was too severe because according to it action can be morally bad even when no bad is intended or foreseen.

There is however also an objection to our proposed definition of moral goodness on the ground that it is too severe and to the definition of moral badness on the ground that it is too lenient. For under certain circumstances is not an intention to do some good enough to warrant the moral goodness of the act even if no good actually results to anybody? And is not similarly an intention to do some bad enough to label the act as morally bad even if no harm actually results?

Accepting these objections thus means that our proposed definitions of morally good and bad acts were at once too lenient and too severe.

I think we must accept these objections—or some of them at least. The proposed definitions cannot be regarded as successful attempts to catch hold of the ‘essence’ of moral goodness and badness. The chief reason why they fail is that they make the moral quality of an act independent of intentions in acting and of the foreseeing of good and harm to other beings.

Before we revise our proposals for defining moral goodness and badness some words must be said on the concept of intending.

8. How is intention in acting related to the foreseeing of consequences?

In order to answer the question we shall have to observe that intention is primarily connected with results of action—and not with consequences. An intention is an intention to do something. That which is intended the object of intention is the result of an act.

Suppose that I open a window with the intention as we say of cooling the room. The cooling of the room is a consequence of the opening of the window. Is not here the object of intention the consequence and not the result of my act? I propose to answer the question as follows:

If I open a window with the intention of cooling the room and the temperature in the room goes down as a consequence of my manipulations with the window then the question ‘What did I do?’ can also be answered by ‘I cooled the room’. Cooling the room is something I can do. There is an act of cooling the room. Its result is that the temperature in the room is now lower than It was before. It is a different act from the act of opening the window to which it has a causal and not an intrinsic relation. The act of cooling the room would be different from the act of opening the window even if opening the window were the only means to cooling the room. They are different because the result of the one is a consequence of the result of the other. But the activity which I display in performing the two acts i.e. the manipulations with the window is the same in both (In this sense the two acts could be said to ‘look’ the same.)

Suppose a person intends to cool the room and with a view to this opens the window. And suppose that he succeeds in making his intention effective. Shall we then say that he also intended to open the window? We would certainly call his act of opening the window intentional. We may wish to call it intended but we may also wish to reserve this term for acts the results of which have the character of ultimate ends in acting. Usage seems to be somewhat vacillating at this point and we need not force ourselves to a decision. But we shall decide to call any act which is done for the sake of or in order to attain some end intentional or intended.

Be it observed in passing that ‘intended’ is used both of the act and of the state of affairs which is its result whereas ‘intentional’ is normally used only of the act. One suggestion could be that the results both of intended and intentional acts be called ‘intended’. I shall adopt this suggestion.

The man who opens the window in order to cool the room can rightly be said to intend two things ‘at once’ or to have two intentions—one to have the window open and another to have the temperature lower.

Although everything which is done for the sake of an end is also intended (as well as the end) it is certainly not the case that everything which is a consequence of action is also intended. If the agent at the time when he is acting does not foresee the consequence (or at least realizes the ‘serious possibility’ that it will happen) then he can in a sense not even be rightly said to have done the consequent thing. ‘Look what you have done!’ we sometimes say of such cases—particularly when we consider the consequences undesirable. This is said in order to draw the agent's attention to a causal connexion of which he was not aware. Once he is aware of the connexion he can not on a new occasion do the first thing without also doing the second.

It is here appropriate to make a distinction between foreseen and rightly foreseen consequences of action. Something is a foreseen consequence if the agent at the time of acting knows or believes that this thing will happen as a consequence of his action. A foreseen consequence will be called rightly foreseen if it actually happens (as a consequence of action). Agents sometimes foresee consequences of their action which never come true.

I shall say that everything which is a rightly foreseen consequence of one's action is also a thing done. But I shall not say that everything which is a foreseen consequence of one's action is also a thing intended. Suppose our man opens a window and that as a consequence the temperature in the room goes down. When asked why he opened the window he answers that he did it in order to hear the birds sing. Let us assume that he foresaw that the temperature was going to sink—or at least was aware of the possibility. Then he could not say that he lowered the temperature unintentionally. But to say that he did not lower the temperature unintentionally and to say that he intentionally lowered it is not—on the ordinary understanding of the words—to say the same. Of everything which is a rightly foreseen consequence of my action I can say truly that I did this and that I did not do it unintentionally. But only of such consequences of my action which were also ends of action—intermediate or ultimate—can I say truly that I intended them or did them intentionally. That consequences are ends entails that they are foreseen—but not that they are rightly foreseen.

This much will have to suffice about the notions of intention and foreseen consequences of action and the related notions of the intentional the unintentional and the not unintentional.

9. Can intentions possess an excellence of their kind i.e. as intentions? Intentions can be firm or vague strong or weak. To say of a man that he had ‘good intentions’ or ‘the best of intentions’ to do a certain thing is it seems sometimes another way of saying that he was firmly determined to do this. We should not ordinarily however call a vague or weak intention to do something a ‘bad intention’.

When ‘good’ is used as an attribute of intentions to indicate firmness of determination or strength of will it can perhaps be said to connote a goodness of its kind. This is one use of the word ‘good’. But it is certainly not that use of the word which we have in mind when we speak of intention as the bearer of moral value or as a component in the moral valuation of acts.

‘Good’ and ‘bad’ as moral attributes of intentions do not connote goodness or badness of its kind i.e. of intentions as intentions. It would seem moreover that ‘good’ and ‘bad’ as moral attributes of intentions are in an important sense secondary. By this I mean that the primary attribution of ‘good’ and ‘bad’ here is not of the intention as such but of the objects of intention the intended results of action. Basically ‘good intention’ is intention to do (some) good and ‘bad (evil) intention’ is intention to do (some) bad or evil. The problem is how to give a satisfactory formulation to the dependence of the value of intention on the value of the intended.

Complications are caused here by the fact that an intention may have several objects and by the fact that everything which is done in order to attain an intended end of action is also intended. Evidently these complications cannot be ignored in an attempt to assess the moral worth of intentions. Be it said in passing that Kant's doctrine of the good will seems to me to suffer badly from Kant's ignoring the ‘multiplicity of intention’ which exists thanks to the fact that human acts have causal relationships of which the agent in acting is seldom totally unaware.

By an intended good I shall understand an end of action which is judged by the agent to be beneficial for some being. By an intended bad I shall understand an end of action which is judged by the agent to be harmful for some being. By a foreseen bad I shall understand a result or foreseen consequence of action which is judged by the agent to be harmful for some being. Every intended bad is also a foreseen bad but not conversely.

An intended good is thus an intended state of affairs the production of which is thought to affect the good of some being favourably. That this state of affairs is produced and that it affects the good of a being favourably are two logically different things. The first may depend upon the agent alone. The second also depends upon the valuation of the being(s) affected by the agent's action. By calling the state of affairs an intended good we thus attribute two objects of intention to the agent. The first is the state of affairs itself. I shall call it his factual object of intention. The second is that some being(s) will value this state of affairs as good for him (them). I shall call this the agent's axiological object of intention.

An intended good or bad can fail to materialize. This can happen for two chief reasons. The intention may fail in regard to its factual object. Or it may fail in regard to its axiological object.

When the intention fails in regard to its factual object we sometimes contrast the ‘good intention’ with the ‘poor performance’. A discrepancy between a professed intention and the actual performance may make us doubt whether the intention was there at all or how firm or strong or serious it was. But if no such doubts occur mere failure with regard to its factual object would not it seems be considered relevant to the question of the moral worth of the intention. One of the reasons for thinking that the moral worth of action resides in the intention alone is probably the idea that this value must not become affected by adversities in acting or by mistakes in foreseeing consequences. This in no way contradicts the view which I think we must in any case accept namely that the moral value of the intention depends upon the value of the thing intended.

There is however also the case to be considered when an intention to do good reaches its factual object but fails in regard to its axiological object. We did something for the sake of promoting our neighbour's good for example and everything went exactly according to our plans—except that our neighbour strongly resented what we did thoroughly disliked it perhaps even badly suffered from it. Should such discrepancy between apparent and real value as we could call it influence our judgment of the goodness of intention?

If by good and bad things we understand things beneficial or harmful as defined in this inquiry then to mistake the apparent value of something which has been done for its real value is to make a false judgment or conjecture either about the consequences of the achieved result of action or about the valuation of this result and its consequences by some subject(s) or about both these things. In a first person judgment ‘I like this’ I would rather have this than be without it’ ‘No thank you’ the subject is valuing something. Then there is no possibility of mistake and no room for a discrepancy between apparent and real value either. In a third person judgment ‘This will do him good’ or ‘This is bad for him’ or ‘He will like it’ we are not valuing but saying what we think that the valuations of subjects are or will be. As far as the value of his acts for others is concerned the agent's judgment will necessarily be a third person judgment and thus not a (genuine) value-judgment. A mistake on his part concerning this value will therefore be a mistake concerning empirical matters of fact.

If we want the moral value of intention to be independent of intellectual mistakes in judging then we must in the name of consistency admit that mistakes as to the value of the factual objects of our intentions do not ‘maculate’ i.e. spoil the goodness of our intentions. ‘He meant it well’ we often say when a mistake as to value (valuation) has occurred and since he meant it thus it may seem illogical to blame him morally for what he did.

But even if on the ground of the goodness of their intentions we morally excuse persons for some evil they have done we nevertheless blame them for ignorance—either of consequences of action or of the beneficial or harmful nature of things or of both. ‘You should have known that this is not the way a man wants to be helped; by what you did you only managed to hurt his feelings’ we say of many a case of intended beneficiality. Because of ignorance much bad is done for good motives. This is an aspect of the matter to which moralists in a Kantian spirit as far as I can see have habitually paid but little attention. Yet it is an important aspect of the moral life—or at least of action affecting the good of man. Paying due attention to this aspect need not however influence our view as to what constitutes the goodness of intention. But it must lead us to realize clearly that intention to do good is by itself of rather limited utilitarian value for the promotion of the welfare of man and of human collectivities.

10. An agent who intends to do good to some being can do so either for the sake of promoting the good of that being or for some other reason. When doing good to somebody is intended solely for the sake of promoting the good of that being I shall say that good is intended for its own sake.

I now propose the following definition of morally good and bad intention in acting:

the intention in acting is morally good if and only if good for somebody is intended for its own sake and harm is not foreseen to follow for anybody from the act; and

the intention in acting is morally bad if and only if harm is foreseen to follow for somebody from the act.

I shall not here further discuss the question whether we may include or must exclude the case that the subject whose good the act is foreseen to affect is the agent himself. (Cf. sect. 7.)

That the intended good must be intended for its own sake if the intention in acting is to be called morally good seems fairly obvious. A master who takes good care of his servants in order that they shall be fitter to work hard for him cannot be said to have a morally good intention in his treatment of his servants. But if from an attitude of gratitude or love he is anxious to see his servants flourish or thrive then we may attribute moral goodness to his intentions.

It seems obvious too that the intention in acting is deserving of moral blame not only on the ground that some harm is intended but already on the ground that some harm is foreseen in acting. Suppose a person opens a window in order to hear the birds sing. Thereby he cools the room and causes another person who is present to catch a cold. He is aware of the fact that opening the window will cool the room and also aware let us assume of the serious possibility that the other person will catch a cold. If ‘with a clean conscience’ he can declare that he opened the window in order to hear the birds and that this was his only end the only thing he intended beside the opening itself of the window then the cooling of the room and the cold which the other person caught were foreseen consequences of his action but not things he can be said to have intended to achieve. Yet it would be reasonable to blame him for having opened the window since he realized at the time of doing it that he was exposing another person to danger. What he intended to do was morally blameworthy because of the foreseen harmful consequences and I see no reason why we should not therefore also say that his intention in acting was morally bad. But we could not rightly say that his intention in acting was malicious. This we can do only when the foreseen harm of his intended action was also intended and not only foreseen.

Some ethicists may be of the opinion that goodness or badness of intention is the only thing that is morally relevant in action and that there is no need to distinguish between the moral value of the intention and the moral value of the act. It seems to me however that a distinction can be sensibly made.

If the intention in acting is morally good but the good which is for its own sake intended fails to materialize—either because the intention fails in regard to its factual or in regard to its axiological object—then the act is not morally good (but morally neutral). If on the other hand the intended good materializes then the act too may be called morally good.

Similarly if the intention in acting is morally bad but the harm which is foreseen does not come about then the act is not morally bad (but morally neutral). If on the other hand the foreseen harm comes about then the act too is morally bad.

It is a noteworthy asymmetry between moral goodness and moral badness that the first presupposes that some good should be intended and moreover intended for its own sake whereas the second only requires that some bad should be foreseen to follow from the act.

It is characteristic of the logical complications of the concepts which we have been discussing that the notion of a morally good or bad act is secondary to the notion of a morally good or bad intention (will) in acting which in its turn is secondary to the notion of a good or bad i.e. beneficial or harmful thing.

11. In our proposed definition of a morally bad act is tacitly presupposed that in every situation there is a course of action open to the agent from which he foresees no harm to anybody. The presupposition in other words is that the agent is always ‘free’ to choose a course of action which is not morally bad.

(This presupposition be it observed is not that there always is a course of action from which no harm actually follows. Harm which follows but which could not have been foreseen at the time of acting is not relevant to the question of the moral quality of the act.)

However the presupposition which we just mentioned is not always fulfilled. Cases may occur in which the agent foresees harm to some being from whatever course of action he chooses. Such cases are not common. But they happen. Their rarity does not nullify the importance which their gravity gives to them.

When harm is (rightly) foreseen to follow from whatever an agent does I shall say that harm is unavoidable. This ‘absolute’ notion of unavoidable harm must not be confused with the ‘relative’ notion of unavoidable harm. By the second I mean harm which cannot be avoided if some particular end of action has to be reached.

It is a logical characteristic of the type of acting-situation which we are now considering that there should be some act open to the agent which is such that bad is foreseen to follow both from doing and from forbearing this act. Forbearing to act it should be observed is also action. Forbearing to act can have consequences just as well as acting can have. The consequences of forbearance is the happening of those things which action would have prevented.

What one would wish to say from a moral point of view of the type of case under discussion is I think the following: In a situation when the agent foresees harm both from doing and from forbearing the very same thing at least one of the two acting or forbearing must be morally excusable. If the harm which is foreseen to follow from doing is less than the harm which is foreseen to follow from forbearing then acting is morally excusable i.e. is not the doing of a morally bad act. If again the harm which is foreseen to follow from forbearing is less then forbearing is morally excusable. If finally the foreseen harm from acting equals the foreseen harm from forbearing either course of action is morally excusable.

We could call this the rule of minimizing unavoidable harm. It determines under which conditions the causing of bad to some being can be morally excused. The harm which it exculpates is unavoidable in the ‘absolute’ sense. Harm which is in the ‘relative’ sense unavoidable can on my view never be morally excused. To argue that it could be excused is a form of arguing that ‘the end justifies the means’. This seems to me the very prototype of immoral argument.

The rule of minimizing unavoidable harm has an obvious plausibility. Yet it takes for granted that a difficult problem can be solved when the rule has to be applied. This is the problem of determining (‘measuring’) the relative magnitude of harm to various beings. It is not certain that this problem has a significant solution in every case. Let us consider the problem in the light of an example.

A man x can save either the man y or the man z but not both from an impending disaster y and z are say wounded and x can carry one of them at a time but not both of them at once to safety. x foresees that if he carries away one of the men the disaster will in the meantime reach the place and consume the remaining man.

We can describe x's possible action in the terms of two choices. His first choice is this: Shall I save one of the two or shall I leave both to perish? If he chooses the second then by his chosen course of action he in a sense becomes responsible for the death or disaster of one man. (x's action is then forbearance.) Or strictly speaking: x becomes responsible for the disaster of one more man than would have perished had x acted differently.

Assume that x chooses to do his best and save one of the two. We would all agree I think that this choice of his is morally right and that a choice to abstain from action would have been morally wrong. To let both perish is to cause more bad than to leave one to perish. The general principle for comparing the relative magnitudes of harm can here be formulated as follows: That all of a number of men suffer some harm is worse (a greater bad) than that some only of these men suffer the same harm. (It is essential that the smaller group of people who suffer harm should be a subclass of the greater group of people and that the harm which each one of them suffers as a member of the first group is the very same harm which each one of them suffers as a member of the second group.)

So far the case seems clear. But having chosen to save one of the two rather than to leave both to suffer x is faced with a second choice: Shall he save y or z? If he chooses the first then as a consequence of his chosen course of action z will suffer. If he chooses the second y will suffer. x we assume foresees this quite clearly. x takes y and carries him to safety. Disaster befalls z.

Can this second choice of x's be justified with reference to the rule of minimizing unavoidable harm? There are two ways in which the justification may be attempted. One is to argue that the harm which follows from leaving z is neither greater nor less than the harm which follows from leaving y and that hence x is ‘morally free’ to choose between saving y or saving z. The other is to argue that more harm follows from leaving y than from leaving z and that x hence is ‘morally bound’ to choose to save y. Both ways of arguing have to cope with similar logical difficulties. Two examples will be given to show what these difficulties could be and how one might try to cope with them:

y is the commander of a group of men of which x and z are members x argues: it is worse for the group to lose y than to lose z. x is then contemplating the consequences of his action from the point of view of the good (welfare) of a group (community) of men. His argument is that it is preferable (better) for the group to have y without z than to have z without y.

y is head of a family z is single x argues: If y is left to suffer or die a number of other people will suffer heavily too. If z is left others will not be seriously affected. Assuming that the bad which y and z are facing is an equal bad for both then it is a greater bad that beside y also others should suffer harm than that no one in addition to z himself should suffer.

The first way of arguing proceeds on the assumption that an appeal to the good of a community of men over and against the good of its individual members is possible. The second argument assumes that it is possible to ‘balance’ the good of two or more beings against one another. Both assumptions were briefly discussed in section 6. Of the second we said that it too is based on the possibility of appealing to the good of a community over and against the good of its individual members. The logical nature and moral character of this appeal however will not be further discussed in the present work. What has been said in this section illustrates the urgency of such discussion.

12. We conclude this chapter with some brief remarks about the use of the word ‘good’ as an attribute of a man.

If there existed a purpose which were essentially associated with man as a kind then one could by the phrase ‘a good man’ understand a human individual who serves this purpose of the kind well. Some may think that there is such an essential purpose of man e.g. that men exist to serve the purposes of their Creator. We shall not here discuss this view.

Men often become accidentally associated with purposes i.e. they are needed and used to serve the purposes of other men or of human institutions. The phrase ‘he is a good man’ is very commonly used to mean that a man fits or serves some such purpose or task well. The man who is thus judged instrumentally good is usually also in some respect technically good viz. as a member of a professional class. Such instrumental or technical goodness of a man however does not mean that he were as a man good even if it is natural to call him a good man.

The idea that there is an activity essentially associated with man as man may be considered inherently more plausible than the idea that there existed some purpose thus associated with him. The function proper to man Aristotle thought1 is activity in accordance with a rational principle or life according to reason. The better a man performs this proper function of his the better he is as a man. This is a clear use of ‘good’ and ‘better’ to attribute to a man a goodness of its kind. I shall not here discuss Aristotle's idea. It seems to me that even if it were true to say that there is an activity which is essential to man e.g. reasoning it would be doubtful whether this activity is of the sort at which a being can be said to be good or bad. To say that a man lives or does not live in accordance with reason is vague but we understand roughly what it means. But to say that a man is ‘good at’ living in accordance with reason would seem to require a special interpretation of ‘good at’ which conceptually distinguishes this case from other cases of that which I have here called technical goodness.

When the phrase ‘a good man’ is used not in the sense of instrumental or technical goodness but with a moral tinge it is related to the notions of doing good and of having good intentions. But it has no clear and distinct relationship to these notions.

A man who is intent on doing good is often called benevolent. A man can be a ‘benefactor’ without being benevolent e.g. if he does good to others mainly for the sake of promoting his own social prestige. A true benefactor must be a man who does good (to others) but he need not be that which is ordinarily called a good man.

I am not suggesting that there is a common and important use of the phrase ‘a good man’ to mean the same as ‘a benevolent man’. But I think it is true to say that when the phrase ‘a good man’ is used with a so-called moral meaning it is related to our idea of a benevolent man. It is of some interest in this connexion to notice that the opposite to a benevolent man i.e. a malevolent man or a man who is intent on doing evil or mischief is quite commonly and naturally also called in a moral sense ‘a bad man’.

One affinity between the morally relevant notion of a good man and the notion of a benevolent man is in any case that both notions have to do with features of human character. A man may do some bad acts and even entertain some evil intentions—and yet be a good man. But the bad he does or intends must count as an occasional aberration. Or it must have some special excuse. If we were asked how much evil the good man can be ‘allowed’ we could of course not answer by giving an exact measure—‘hereunto and no further’. But we could give an inexact and yet significant measure by saying that the bad he does or intends must not affect our judgment of his character. A good man may do some mischief or revenge a wrong which he has suffered from another man or tell a lie. But he cannot be mischievous or revengeful or untruthful.

With this last remark we are also touching upon one of the differences between goodness and benevolence as attributes of men. A benevolent man is not necessarily a virtuous man and he may be lacking in a sense of justice. Virtue and justice are two prominent features in our picture of moral excellence. Until we have discussed them our notion of the good man will remain insufficiently clarified.

• 1. See especially the discussion in EN, Bk. I, Ch. 7.