by Ted Honderich

In the days after 11 September 2001 the awfulness of what happened seemed to make impossible any detached reflection on it, let alone what would be taken as anti-American reflection. In my own case this came to be followed, maybe too soon, by a quieter recollection of other things and also a sense of obligation. 

The recollection has to do with a comparison between the worst-off tenth of population in the poorer world and the best-off tenth in the richer world. The former tenth has more than 40 years less of life than the latter tenth. It is as if these millions of human beings were a different species. This is not the only relevant recollection, and perhaps not the most relevant. The most relevant may have to do with one people being dispossessed of their land and degraded. I refer to the Palestinians. 

The obligation is simple and also has two parts. The first is the necessity of remembering all victims, not only those in New York and Washington. They all have faces and histories and people who cry for them, even if they are not like us and their photographs appear in no newspapers and on no television screens. 

The second part of the obligation is to say plainly what seems true. One's present sense of this necessity can be owed in part exactly to the awfulness of 11 September 2001 rather than impeded by that awfulness. More precisely  the necessity is to say something true that enters into the explanation of that day and will enter into the explanation of any more such days.. 

The piece that follows here was written long before that day, but it gives a view of mine of it. It follows on from Terrorism and Philosophy 1: On Inequality and Terrorism, and Differences We Make Between Them.

I have not changed anything in the piece, which is about the distinction between acts and omissions. If changes had been made, they would include my newer conviction that what the rich world, America above all, does to the poor world, is now not so much a matter of omissions as it was and is instead more a matter of acts

We and the Americans are now less excused by the fact that what we and they do can be less than fully intentional. We now know more of the far greater awfulness, the immeasurably greater awfulness, of what we are doing.


Is there not all the distance in the world between our lives and the lives of the violent? Is there not a difference in moral kind between our conduct and theirs? Almost all of us believe something like this. We believe that our ordinary conduct in life is right, or right on the whole, or right enough, and thus far away from the conduct of those who kill, maim, and wreck for what they take to be purposes of justice. That is wrong, worse than wrong. The belief that I have in mind is about kinds of conduct, not kinds of people. 

No doubt almost all of us are also confident that we are better persons, morally better, but what I have in mind is a contrast between acts as distinct from agents. Acts are a related but different subject-matter, as indicated by the simple truths that one can do the right thing for a bad or shameful reason, a reason that lowers or ruins one’s standing as a person, perhaps marks one as inhumane, and one can also do the wrong action for a reason or from a feeling that does one credit.

Most of those who engage in political violence of the Left do not agree with our view of their conduct and ours. If we look to what they say in disagreement, what we find may be bound up with a Marxist or another ideology, or with some history of wrongs done to a class or a people or a group of peoples. From each of these things, however, it is usually possible to take away a general proposition that can be considered on its own, free of certain doubtful presuppositions, of obscurities, encumbering terminologies, and too much of the past. It is that we who are law-abiding, contrary to our common belief, do not live lives which are anything like right on the whole. Our conduct is wrong.

The argument for this is that by our ordinary lives we contribute to certain terrible circumstances. We make essential contributions to the shortening of the lifetimes of whole peoples and classes, and to many kinds of suffering and distress and degradation, and to denials of autonomy and of freedoms. In fact we ensure by our ordinary lives that multitudes of individuals die before time, that families exist in single wretched rooms, and that this or that people are powerless in their homeland, or subservient in it, or driven from it.

Of course the judgement by the violent that our conduct is wrong will be taken as in several ways mistaken, no doubt as absurd. At best, it will be said, it is tu quoque, a retort by the guilty that their accusers do wrong. As with tu quoques generally, it will get a scornful reply, to the effect that the guilty are trying to divert attention from their conduct, trying to change the subject, perhaps a subject which finally they cannot face. 

This is too quick. There is a general question about such tu quoque retorts and such replies to them. If a tu quoque can be established as true, is it none the less to be ruled out of bounds? More particularly, if it can be shown that we all do wrong in our ordinary lives, can this be put aside? Can it be shown that nothing follows about the conduct of the violent? It may be, rather, that certain things are somehow inseparable, in logic or in fact. I shall now take this to be true, and in the end say more about it.

There is the prior question, certainly, of whether it actually is true that we do wrong in our ordinary lives. Of course there are replies to this. A principal one of them, which is my subject in this essay, is that our ordinary conduct, even if it does contribute to terrible circumstances, consists in no more than omission. 

It is not as if we acted to shorten average lifetimes or to secure other terrible ends. That, of course, would be wrong. It may be argued in support of this that there is a certain difference of fact between every act and the related omission, a general difference of fact which issues in a difference of morality. In virtue of the general difference an omission may be right or at any rate less wrong when a related act would be wrong or more wrong. It may be argued alternatively that there is a limited difference. Some omissions, including the ones in question, differ from the related acts in a factual way, with the given difference having similar moral effects, similar effects with respect to the rightness or wrongness of actions.

But (a) is there really a morally important difference of fact between every act and the related omission? and (b) Is there a difference of fact between some acts and their related omissions, which difference gives rise to a difference in rightness? Rooted beliefs are in question here, but a small number of moral philosophers (1) have nonetheless been inclined to doubt or deny that there is a general difference that will do the required work. Some have also doubted or denied that there is any difference between even some acts and their omissions that makes for a difference in rightness in those cases.

My intentions in this essay are to deal with these two general issues about acts and omissions, and so to arrive at or at any rate come closer to a decision on the judgement of the violent that we in our ordinary lives do wrong. The two mentioned issues about acts and omissions are a subject which is wider than that of political violence, one which has not often been discussed in relation to political violence and perhaps never at all in what can be called a non-ideological way. It does not begin to follow, of course, that the issues are not relevant to violence.

It is true that acts and omissions is a subject in itself. To anyone who feels a reluctance to consider one whole subject in the context of another, it may be worth remarking that in political philosophy as elsewhere one can keep one’s nose too much to the grindstone. There is a kind of concentration which defeats its own end, or at any rate does not advance its own end, the end of getting a clear view of something or of fixing its true weight. 

Given our concern with political violence, we might spend time not only on the general question of acts and omissions, but also on war, and on punishment by the state, and on other things. Their relevance to political violence is of several kinds, and it is considerable. If it is true, as it seems to be, that most of the wars of the past ought not to have been fought, and if we were to come to have an understanding of their wrongfulness, we would be more able to judge political violence. So too with punishment by the state, whatever its recommendation or want of recommendation.


The conceivable act of sending poisoned food parcels to India is not one that it would cross our minds to perform. However, how is this possible act related to our omitting to give what money we could to Oxfam, or to some other starvation-relief organization? The result of our omission is that some people die. (2)

We would not knowingly vote for a political party which aimed to secure that the economically worst-off tenth of the population in European and other economically developed societies continues to have an average of about five years less time to live, whatever the party had to say about the deserts of these people and their betters. How does the conceivable act of voting for such a heinous party stand to our omission in not really working to try to change these average lifetimes?

We would not contemplate infecting others with serious diseases, but we do not try to secure, as we could, that people have equal and more than minimal medical care. We do not try to secure that end for people in our own societies, let alone people generally.

We would not remove library books from poor schools, whose pupils can have but the smallest hopes in life. Most of us would not try to increase our private profits by evicting families and hence making them homeless. But we do not work as we might for equal and good education, and we do not work for an end to homelessness.

We would not contribute funds to a dictatorship, or to a minority engaged in denying autonomy to a majority. At any rate most of us would not. However, we do in fact fail to contribute funds to political movements against dictatorships, and to political movements against oppressive minorities.

Some may say that there are more ‘realistic’ ways of dealing with starvation than by contributing to Oxfam, and more realistic ways than those mentioned or implied of dealing with curtailed lives, ill-health and disease, bad education, homelessness, and oppression. Suppose that there are better ways, perhaps joining this or that political party or forwarding some sort of economic policy, perhaps one of ‘self-help’. This will not matter much to our inquiry. Just as we do not contribute what money we could to Oxfam, we do not do these other things either, whatever is to be said of their worth. We omit these too, thereby as quickly raising the problem of omission.


In what ways are acts and omissions the same and different? Let us consider similarities and differences, differences which hold between all acts and their related omissions and also differences which hold between only some of them.

(i) An act is a bodily movement or stillness deriving from intention, or a sequence of movements or of movements and stilinesses. To say this is to move with some speed past a number of philosophical questions, one of them about the nature of the mentioned connection between intention and movement, the connection described by saying the second derives from the first. Still, it surely is no longer the case that the given definition of an act is controversial. Certainly the definition does not make an act into just a movement or stillness. Rather, an act is a movement or stillness with a certain mental ancestry.

What is an omission? One thing to notice is that the term ‘omission’ can be used in a number of ways. To mention two, it can be used for (a) the omitted action, or what was omitted or not done, and also (b) for the omitting action, or what was done instead. 

If I omit to turn off the radio at 12 noon, we can use the term ‘omission’ for what I did not do: turn off the radio at 12 noon. An omission in this sense might be described as an unrealized possibility, or indeed as a mere nothing. Perhaps this conception has had something to do with rather vague scepticism about omissions being morally important. How could a mere nothing be wrong? Of course, this could not be the only conception we have of an omission. Not all omissions could be mere nothings. This is so, to mention but one reason, since we do often witness what we call omissions. These, obviously, are omissions in the second sense, exemplified in the example not by my turning off the radio at noon but rather by my not turning off the radio at noon. An omission in the second sense, which is the more important one, and the one with which we shall be concerned, is clearly not an unrealized possibility.

How did I not turn off the radio at noon? As it happened, I did not turn off the radio at noon by staying my hand, by performing that little action after having made a first move towards turning off the radio. In staying my hand, further, I did not turn off the radio. We can correctly say more than that it was by staying my hand and in staying my hand that I did not turn off the radio at noon. My staying my hand was or was identical with my not turning off the radio at noon. What we come to, then, is the idea that an omission, in the most important sense of the word, is an action, or, as we may add, a sequence of actions. My not going to the West End over the weekend was my staying in Hampstead.

We have not got a difference between all omissions in this sense and their related acts, obviously, but a sameness instead. Omissions like acts are bodily movements or stillnesses deriving from intention, or sequences of such movements and stillnesses. We might well refer to omissions and acts as negative and positive actions. I shall in fact stick to the familiar terms, ‘act’ and ‘omission’. (I shall use the term ‘action’, by the way, as against ‘act’, for either an act or an omission.) We shall of course see more of what leads us to regard two actions as being an act and omission, as related in that way. One main thing is that there is the same or a similar or a connected upshot on the whole. Another, is that when either the act or the omission is done, the other was possible.

There is of course a great difference between an act, taken as something actual, and an omission in the sense of the ‘omitted action’ since the latter is not something actual. It is important not to run omissions of the two kinds together. It is possible to do so, and to become a bit mystified, partly because an omission in either sense brings in an omission in the other. If there was an omitted action, there must have been the omitting action, and if there was an omitting action, there was what was not done. It is not necessary, but it will be simplest for us to proceed in terms of omissions in the sense of ‘omitting actions’. These, to repeat our first proposition, are like acts in being certain bodily movements or stillnesses, or more likely sequences of these.

(ii) Are all omissions simply actions described in a negative way, actions with descriptions which include a ‘not’ or something like one, while the related acts are described affirmatively? The trouble with this simple idea is that we do not actually get a division into acts and omissions, since each and every action falls under both affirmative and negative descriptions. My not turning off the radio was also my staying of my hand. 

Another idea is this one: an omission is an action or a sequence of actions which falls under a negative description forthe agent, while the related possible act would fall under an affirmative description for the agent. That is, the agent describes the omission negatively and would describe the act affirmatively, whatever other people do or would do. This is so because his intention was what it was. We can imagine a doctor, for example, who carries on in a certain way, which does not include giving penicillin to a badly deformed infant to save him from pneumonia, and describes his action as ‘not trying to keep the infant alive’. He would describe a related action, perhaps the giving of a lethal injection, as ‘killing the infant’.

This common idea is no better at giving us a general distinction between acts and omissions. It does not give us a difference between every act and the related omission. Consider another doctor who also carries on in such a way that a person dies. He too does not give penicillin for pneumonia. However, he describes his behaviour as ‘letting the person die’, or ‘freeing an incurable patient from suffering’, or ‘doing my duty’. One of these affirmative descriptions is chosen by him. He may even refuse, for whatever reason, to describe his behaviour as ‘not trying to keep the person alive’, and also refuse to use any other negative description. This is certainly possible, however unlikely. The plain fact is that this would not stop us from regarding his action, when compared with giving a lethal injection, as an omission. We here have an omission and a related act such that the omission is not negatively described by the agent.

If we return to the six examples given above, we find the same thing. If we compare sending a poisoned food parcel to India and omitting to contribute to Oxfam, we shall take the first to be an act and the second the omission even if the agent prefers to describe the second act affirmatively, as ‘buying wine’ or ‘saving money’ or ‘looking after my own children’. This is not to say anything of the rightness or wrongness of the actions, of course. In each of the other five examples, the same is true. No matter the description of the agent, the second act of the pair counts as the omission, for reasons to which we will come.

Still, there is something else to be said here. We all have a tendency to describe the acts affirmatively and the omissions negatively. This is the natural thing for both the agents and others, evident enough in what has been said so far in this essay. For what it is worth, then, we here have a difference of fact, our first one, but one between most rather than all acts and omissions.

What is the explanation of this fact that in the case of most acts and omissions, the omissions are described negatively and the acts affirmatively, by both the agent and others? The question is of interest to us for several reasons. It may be suspected that there is some general difference of fact concerning acts and omissions which is a part of the explanation of the linguistic tendency, which general difference of fact may give rise to a difference in rightness. (The idea would be that the general difference of fact, for some reason, does not always give rise to negative description of the omission and affirmative description of the act, but it does always affect rightness.) Again, it may be suspected there is a limited difference of fact, concerning some acts and omissions, which difference wholly explains our linguistic tendency and gives rise to a difference in rightness in the case of the given acts and omissions.

There is of course an explanation of our linguistic tendency, but, as may be anticipated, it has to do with us rather than with acts and omissions in themselves. That is, it does not have to do with any fundamental properties of some actions but not others, which properties are such as to call out for the word ‘not’ in their description.

Notice first that it does seem to make sense to say that there are differences between things as affirmatively described and things as negatively described. An affirmative description is such that a thing so described has a property and a negative description is such that a thing so described lacks a property. An affirmative description, secondly, brings in a single property such as redness, while a negative description brings in a collection of properties, those which are other than redness. 

Whatever may be thought of these claims, which are a part of the traditional philosophical problem of negation, they do not get us the conclusion that it is a difference in fundamental properties of acts and omissions that explains our tendency to describe acts affirmatively and omissions negatively. This is so, to repeat an essential point, because any action which falls under an affirmative description also falls under a negative description. In every case there are true affirmative and negative descriptions, whether or not we are inclined to use one sort more than the other. Of every action without exception, then, we can say that it has properties of a kind associated with both affirmative and negative description. The explanation of our linguistic tendency does not have to do with there being two separate lots of things, one with one sort of fundamental property only, and the other with the other. All of the things have both sorts.

The true explanation of our linguistic tendency must be otherwise. Let us speculate about what may be part of it. This has to do with an established habit which we carry over from a significantly different kind of situation, but one in which we also talk of what we call omissions. It would in any case have been a good idea to distinguish and set aside this kind of situation and usage. Not to do so would have left open a door to confusion.

In this situation there is no pairing of actions of the kind we have been noticing. To begin with, we speak of an omission but there is no related act: no act with an identical or similar or connected upshot on the whole. On the contrary, there is an opposite upshot. In this kind of situation, secondly, it is more or less generally accepted that the agent had a particular duty to do what he did not do. There is a generally accepted rule, perhaps a law in the ordinary sense. Obviously this is different from the sort of situation and talk which is our main subject-matter. The duty and the rule or law, thirdly, pertain to an individual’s special position or role. For this reason, by the way, we can call the omissions in question role-omissions.

Consider a judge, who has a duty to give a certain instruction to a jury, at a fixed point in a trial, so that they will not be prejudiced against the defendant in a certain way. This is specified by law. Suppose the judge says nothing at the given point. We call this an omission, certainly. However, we do not have in mind a certain conceivable act with an identical, similar or connected upshot on the whole. If we did, it would be more or less this: the act of instructing the jury to have or to act on a certain prejudice against the defendant. (We in fact have in mind the action of so instructing the jury as to try to prevent prejudice.) We think the judge’s omission is wrong not because we have compared it with the conceivable act but because the judge went against his duty, fixed by law, in his act of omission.

It is our established habit to describe an omission of this sort, a role-omission, in a negative way. This is so, presumably, because there is a settled rule which specifies that a person is to do precisely such-and-such. His error, according to the rule, is most precisely or carefully described as not doing such-and-such, and some care and precision are natural in important and more or less formal circumstances. More might be said. Whatever the reason, however, there is the fact of our established habit with role-omissions. To come to the main point, it is fairly certain that it is part of the explanation of our tendency to use negative descriptions for the other omissions, those with acts related to them in the way we know. ‘Nots’ are standard in connection with role-omissions, and so come to be used with the other sort of omissions.

(iii) Is the causal connection between an omission and the result different from the connection that would hold between the conceivable act and the result? It is sometimes supposed that it is very different indeed and significantly so. It has been said that the result, perhaps a harm, would be the result of one’s movements in the case of an act, but, in the case of the omission, it is the result of conditions which one omitted to change. These words are not terribly clear, but they manage to suggest that the act would be more efficacious than the omission with respect to the result. An act would somehow be more of a cause than the omission. 

There is the same suggestion in the related statement that the great difference between an act of mine and an omission of mine is that the result of the omission would have occurred even if I had never been born, but that is not true of the act. The same children who are now deprived of good and equal education would also be deprived of it if I had never existed. On the other hand, the result of my conceivable act of removing library books from poor schools would of course depend on my existence.

Certainly there is a truth in this last point, of whatever significance. One way in which it can fail to happen that I do something is by way of my not existing at all. It is plainly false, however, that my failing to do something, a particular action, is in any way less efficacious with respect to a result than would be my doing something, another action. The omission is precisely as much of a cause as the act would be. Staying my hand at noon, my not working for good and equal education, each does obviously make a difference. It will be as well, however, to have some details of the causal nature of acts and omissions.

An act in almost every instance would be a necessary condition, in the situation, of a result. That is, the result would not happen in the situation without the act. Also, the act would be part of a causal circumstance, a set of things sufficient to produce the result. That is, it would be part of something which necessitates the result, which would not obtain without being followed by the result. Indeed, the act would complete the causal circumstance: the other things would already be on hand. Sending a poisoned food parcel would be a necessary condition, in a certain situation, of certain deaths: the deaths would not happen without the parcel. The act would be a part of a causal circumstance which was sufficient for the deaths. It would complete it.

We can say near enough the same things of an omission of mine, say the action by which I got rid of 500. Suppose I bought a chaise longue. It would be mistaken or misleading to suppose that doing exactly that was a necessary condition of the death of a starving person. Still we can suppose that a necessary condition of the death was that in one way or another I omitted to contribute 500 to Oxfam. By my action I did supply the necessary condition. That fact is not put in question by the fact that I might have supplied it another way. Furthermore, my omission was part of a causal circumstance sufficient for the death. It completed that circumstance. 

Given all this, it is certainly mistaken to think that the act of sending poisoned food would be more efficacious, more of a cause, than the omission. Similar comments, although not exactly the same ones, are in place with our other five examples of act and omission.

It is clear enough that very often the causal circumstance in connection with omission and result is in another way different from the circumstance in connection with act and result. In the case of the omission, the rest of the causal circumstance for the result is likely to include the omissions of other people, perhaps many other people. Consider the result which is the particular loss which may be said to result from my failing to work for equal and more than minimal medical care. The omissions of many other people are also parts of the causal circumstance for that result. If we turn to the conceivable act, my intentionally infecting people, there is at least the possibility that my act would be the only action in the causal circumstance for the result. Related remarks apply to all but the second of our six examples, the one having to do with voting for the appalling political party.

It would also be misleading to describe this difference in connections by saying the omissions usually are lesser causes than the related acts would be. They have much the same causal features, those enumerated a moment ago. There is only the difference here that it is usually true that actions of others, perhaps many others, enter into the causal circumstance for the result in the case of omissions, and this would not be so in the case of the related acts. There is the related small fact that it is generally true that if I act, intending to cause a harm, there is a small possibility for others to prevent the harm, but ill omit to act, thereby not myself preventing the harm, there is a larger possibility. I leave more scope for good works and self-improvement by others if I omit to act. There is not much to be made of this, of course, unless one brings in certain relgious ideas which are now of the past, and which surely were never sufficient to their tasks.

There may also be other differences having to do with causation between acts and omissions. Often but not always the causal chain or causal sequence in the case of acts would be shorter or ‘simpler’ than in the case of omissions. Clearly enough a causal circumstance is not less efficacious simply because its linkage to the upshot is of either of these kinds. It is true that certain other features sometimes go with shorter as against longer and ‘simpler’ and as against ‘more complex’ causal chains, and we shall notice these features in due course.

(iv) If we consider omissions from the point of view of the intentions of the agents, we can make a very rough division into three categories. It has to do, in part, with the fact that any action, as an omission, has what we can call a certain nature or character and a certain effect. One of my omissions is ‘not contributing money to political movements against dictatorships’. That description gives the nature of the omission. The effect is that the given political movements are to a degree weaker. It is the agent’s relation to the character or the effect of his action that determines, as we can say, whether his omission was fully intentional, partly intentional, or unintentional.

Fully intentional omissions are taken to occur in medical contexts. Consider the doctor who continues his ward-round during a certain time. To do this is to perform a certain omission, failing to give penicillin to a patient for pneumonia. This has as its effect the death of the patient. If the character of his actions or their effect, as we can say, was the intention of the doctor, then the omission was fully intentional. To put it differently, if the actions derived from an intention which the doctor would characterize to himself as the intention not to give penicillin to the patient for pneumonia, or the intention to bring about the death of the patient, then his actions were a fully intentional omission.

It is worth noting that the intentions that issue in fully intentional omissions are in all respects like the intentions that issue in fully intentional acts. Some are impressive, as we might say, since they involve internal resolutions or declarations to do such-and-such, and some are not. Some are such as to take up or occupy the whole attention of their agents, and some are not. Some are formed under higher-order intentions and some are not. Some are, so to speak, something like the whole story of what the agent is thinking and feeling, and some, no doubt most, are not.

Partly intentional omissions are connected with earlier actions and their intentions. The earlier intentions are more important than the immediate intentions of the omissions, those that just precede or accompany the omissions. Out of bloodymindedness, I do not write down a note of an unpromising engagement when asked. I have it in mind that I may forget. There is the result, given my general forgetfulness, that I do forget and do not turn up. My not turning up is one example of a partly intentional omission. Here is another. I look away from the Oxfam advert in my newspaper, including the photo of the starving child, my intention being not to be troubled or got at by it, and indeed not to have to face a moral challenge and to give money. There is the result of this action and a multitude of related ones, that I later buy wine without thinking that doing so is in fact not contributing to Oxfam. 

A partly intentional omission, then, does not in fact derive from an immediate intention which specifies its character or effect, for the reason, roughly speaking, that the agent has so acted earlier that it will not. A partly intentional omission does derive in part from an earlier intention which does have in it something of the character or effect of the later action. The earlier intention has something of the later omission in it. Needless to say, the characterization is very vague. Needless to say, as well, there are many sorts and varieties of what have been labelled partly intentional actions. In fact, my main argument has not much need of greater detail, but I shall say a bit more later in other connections.

Unintentional omissions, finally, are those whose characters and effects are not intended by the agent, and it is not the case either that this is so because of earlier activity of his. He did not so act or plan or contrive at an earlier time to have the omission out of mind when he performed it. If the idea of an unintentional omission comes to anyone as a surprise, this may partly be because of the fact that there is a tendency to think of all ‘omissions’ as blameable. There is also the fact that role-omissions, as they were earlier called, are very unlikely to be unintentional and are pretty well always blameable. As we are conceiving omissions, and as they will be defined in due course, it is possible that they be unintentional. Nothing much hangs on this decision, however.

This three-part division of intentions certainly does not apply to the acts which we compare with omissions. This is a considerable difference between act and omission. This may also come as a surprise, and indeed the surprise derives from a truth. If one takes any of the acts in our examples, say removing library books from poor schools, it is possible at least to imagine a related partly intentional action, and a related unintentional action. That is one fact. Another, however, is that our comparisons of act and omission invariably involve only fully intentional acts. That is our practice, how we go on. One difference between acts and omissions as we have them, then, is that the conceivable acts are fully intentional ones, and the omissions may be fully intentional, partly intentional or unintentional.

I trust there will be no suspicion that this distinction between acts and omissions is somehow arbitrary, or that it will have the effect of begging some question. It is a distinction between acts and omissions as we have them, between the actions which we relate as act and omission. In fact, as will become apparent, my conclusions do not depend greatly on the fact that all acts, unlike most of their related omissions, are fully intentional. More precisely, it is not greatly important to my conclusions that there are possible actions which in fact we do not consider but which stand to acts as partly intentional and unintentional omissions stand to fully intentional omissions

(v) There is a related way in which some acts and omissions differ, having to do with the natures of their intentions. The intentions that would issue in the conceivable acts in the six examples are horrific or insane, inhuman, vicious, grasping, uncaring in the extreme. Let us give a moment or two to details of this intentionality. The agent, in the case of any of the conceivable acts, would be aware that there were victims of his act, persons at least badly affected by it. He would be aware that victims existed. In the case of some of the conceivable acts, the agent’s awareness would derive from the fact that he would have to bring himself into the neighbourhood of the person or persons affected by his act. In the case of other acts the agent would necessarily have a lively awareness of the existence of victims without actually coming into their neighbourhood. This is true of the sending of the food parcel and, in different ways, of voting for the appalling party, evicting people for greater profits and contributing funds to a dictatorship.

It is worth noticing separately that in the conceivable acts the victims would be individuated by the agent. One way in which a victim is individuated, we can say, is by the agent knowing such individual features of the victim that the victim, for the agent, is a person. We might spend time making this notion more precise but it is not necessary. We can also allow another way of individuation, whereby groups of victims are known to the agent by shared features, perhaps connected with membership of a social class, which again makes them persons for the agent. As will be plain enough, individuation is something over and above just the awareness that victims do exist. There would be individuation in the case of all our six conceivable acts.

Let us compare this with the natures of the intentions in the case of the omissions or, more precisely, in the three cases of omissions. We must obviously distinguish between unintentional, partly intentional and fully intentional omissions.

Unintentional omissions, as we know, are actions such that neither the agent’s immediate nor his earlier intentions involved the character or effect of his action taken as an omission. His intentions, then, are in no way suggested by this action taken as an omission. His immediate intention will have to do with his action taken as something other than the omission, perhaps buying a car or putting his money into a building society. It may be an intention of an ordinary or of course a good kind. In unintentional omission, there is no awareness of the existence of victims, no individuation of them.

To pass on to the nature of fully intentional omissions, there is unlikely to be an instance of any of the omissions in our six examples being performed in a fully intentional way. This is so, at least, if we take certain more precise versions of the examples. 

It is unlikely to happen that because of some extraordinary situation, a man fully believes that his omitting to contribute a certain sum to Oxfam on a particular day will result in the deaths of several people known to him, and he does do something or other rather than contribute, and he does this in the desire that the people die. That is, he has the intention of their deaths in the way that the related intention is had by the conceivable man who does the act of sending a poisoned food parcel. The omitter may not have it in mind to support his family or to do any such thing by his action. Such an omission would in fact derive from an intention about as horrific as the intention in the case of the related act. There would also be a similarity of intention in our other five examples. That is, the intention in the omission if it were fully intentional would be bad, and about as bad as the intention in the conceivable act.

There is a traditional problem about the nature of the intentions in partly intentional omissions, and the goodness or badness or whatever else of the agents. What has been said here so far, not very usefully, is that the omissions derive in part from earlier intentions, which intentions have in them something of the character or effect of the omissions, in our examples a black character and effect.

We rightly feel that there is more, perhaps a good deal more, to be said against the man whose failure to contribute to Oxfam is partly intentional, as against the man whose failure is unintentional. There is, it seems, a very great deal less to be said against the man whose failure is partly intentional, as against the man whose failure is fully intentional. Still, how exactly are we to judge partly intentional omission? Reflection on the question leads in several directions. 

It does seem settled that there is a great difference between the earlier intentions and activity in the case of partly intentional omission and the intentions in either the related fully intentional omissions or of course the related acts. There is nothing less than a gap between not reading a newspaper account of a famine, albeit out of a certain selfish impulse to avoid the matter, and sending a poisoned food parcel or performing a certain conceivable omission with full intention. One can say that almost all partly intentional omissions reflect better or less badly on the agents than the related fully intentional omissions or the acts. This is partly a matter of the degree of awareness in the earlier intention of the existence of victims of the possible future omission. There is also a point to be made about individuation.

We shall later spend some more time on the subject of partly intentional omissions. For now, we have the comparison that in many cases of act and omission, the conceivable act would derive from an appalling or at least bad intention, while the omission if it is partly intentional derives from intentions which do not throw so bad a light on the agent. The comparisons with the fully intentional and the unintentional omissions are of less interest.

It is to be remembered, by the way, that not all cases of act and omission are like the six with which we began, where the intention in the act is as bad as in fully intentional omission, worse than in partly intentional and unintentional omission. Consider the three possible versions of an omission of a kind mentioned above, a doctor’s not giving penicillin to an old, dying, suffering, and incurable person. It may be that neither the omission if fully intentional, nor the related act of killing, comes from anything like a bad intention. Partly intentional omission would here be worse and indeed unthinkable, and unintentional omission worse and no doubt culpable. There are also other conceivable cases of act and omission which are different from our six.

(vi) Let us turn from intentions and agents to customary morality, to ordinary judgements about the rightness or wrongness of actions. In the case not of all but of many acts, including all those in our six examples, there is a general acceptance that the act would be wrong, or worse than just wrong. There would also be a general acceptance that the related fully intentional omissions would be wrong. The same thing, whatever we are to make of the fact, is not true of the related partly intentional omissions.

The acts and the fully intentional omissions would be taken to offend against one or more rules, principles or laws. The act of contributing to a dictatorship would offend against fairly ordinary sorts of political principles, and no doubt against international declarations of human rights. So with the related fully intentional omission. The same and more is to be said of the conceivable act of voting for the appalling political party with the intention of keeping the lives of some people shorter. So with the related fully intentional omission. Removing library books from poor schools, and evicting families for private profit, and the related omissions, would offend at least against standards of ordinary decency or humanity. Infecting people would go against both ordinary morality and law, as with the related omission. The act of sending poisoned food parcels would outrage anything recognizable as a morality, and also be against ordinary law. So with the related fully intentional omission.

Whatever view we ourselves in this inquiry come to have of the rightness or wrongness of the related partly intentional omissions, then, they are not now accepted as wrong. It is not as if they go against ordinary political principles, or standards of ordinary decency, or ordinary morality, or ordinary law. What I have in mind is a matter of the non-existence of certain conceivable attitudes, practices, and institutions, just as it is a matter of fact that the related actions and fully intentional omissions are accepted as somehow prohibited. Far from being prohibited, many of the omissions may well be accepted as required or obligatory. If I were to increase my contribution to Oxfam to the point where my children lived poorly, in comparison with certain other children in my society, I would run afoul of an established outlook. That outlook requires that I fulfil certain ordinary obligations, that I have and act on certain loyalties, and, in order to do such things, that I omit various other things.

There remain the unintentional versions of our six omissions. Ordinary morality obviously does not take them to be wrong. They are hardly distinguished from the partly intentional omissions.


The following comparisons between act and omission, as we shall see, have another importance.

(vii) Acts and omissions give rise to what we have so far called results or effects, and might more precisely call main effects. These are usually the events or states mentioned or implied in the description of the action. In the six examples, they are the deaths of some people, shorter average lifetimes of some people, the states of health of certain people, poor education of certain children, homelessness of some families, and political oppression.

In general the main effects in question with acts and omissions are harms and benefits. Harms so-called are to be taken as including awfulnesses and catastrophes. The problem we are considering arises mainly from the fact that some harms arise from omissions, and so raise a question about whether the omissions are as wrong as certain terrible acts which would have similar or identical, similar or connected effects. There is also the problem, not illustrated by our six examples, that some benefits are as much the issue of terrible acts as they are of omissions, and so raise the question of whether the acts are right. The cases of killing or letting-die are central here.

Not much needs to be said of the nature of harms and benefits. It can be assumed that there is some fairly wide agreement about what states of affairs are to be avoided or prevented, if we can, and what states of affairs are to be secured, if we can. Such an agreement was assumed, of course, in my listing a moment ago of the harms involved in the six examples. There is a considerable problem, certainly, of what general and systematic summary can be given of such events or states. The answer which I favour is that benefits are states or events for which there is something to be said if one is committed to the well-being of everyone without exception, and to something of a related kind for other sentient things. Harms are characterized in terms of conflict with the same ideal. We can agree for the most part on what things are harms and what things are benefits without agreeing on the general systematic summary.

The principal point to be made about main effects has already been in view. It is that the main effect of an act may or may not be the very same as the main effect of the related omission. A comparison as to effects may produce a sameness or a difference between an act and the related omission. It is more likely that there will be a difference. This remains so even when the effect of the act can be described in the same words as the effect of the omission. Generally, of course, there is some similarity between the two main effects, which similarity leads to the two actions being brought together as act and omission. Occasionally, there is connection in main effect, not identity or similarity, as in a situation where a man’s act will be the killing of one man and his omission will for some reason result in someone else’s killing of twenty. (3)

Main effects, to repeat, are states or events mentioned or implied in the descriptions of actions. They are not to be taken as including effects on agents themselves unless this is specified. Nor do main effects include what may be called side-effects. We shall notice both these things separately.

(viii) Some acts make it more or less certain that main effects, whether harms or benefits, will occur. So with some omissions. For the most part, however, there is less than certainty. There is only probability, something in between zero and one. The principal fact to be noted, however, is that it may or may not be true that the probability of a main effect, given an act, is the same as the probability of a main effect, given the related omission. Usually the probability of the two possible effects is not the same. If the probability of an effect given an action is sometimes close enough to be indistinguishable from the probability of a similar or identical effect given the omission, the ordinary case is otherwise. A moment’s attention to the examples indicates this. Again, it is a simple but important fact.

(ix) Connected with things already noticed, but distinct from them, are what we can call experiential effects on the agent. An act or an omission is within the experience of the agent, and hence has effects on him at the time and thereafter. These are likely to be different as between act and omission. The different emotional concomitants of killing as against letting-die may come to mind. One relevant fact in connection with the difference in experiential effects has to do with responses already mentioned having to do with attack and the like. However rational it may be, we have a livelier experience of attacking than of not-saving. 

A larger relevant fact is that our customary moralities, as we know, include prohibitions of many acts, and so give rise to accusation and self-accusation. What we feel when we remember our partly intentional omissions is different from what we would feel if we were to remember that we had performed the related acts. The difference in experiential effects, of course, will be different in the case of fully intentional as against partly intentional omissions, and so on.

(x) In addition to main effects, and experiential effects on agents, there are side-effects. These may be effects on victims, other than main effects, or they may be effects on third parties. Any of our examples can be used to illustrate the fact that the side-effects of an omission are likely to be different from the side-effects of a possible act. There are, for example, the side-effects on others in my life if I increase very greatly my contribution to Oxfam. These would be different, certainly, from side-effects of the conceivable act. In the second example there would be the side-effects on the economically worst-off tenths of there being a political party with the aim of keeping their average lifetimes shorter. There are possible economic side-effects to be taken into account with respect to our ordinary activities in omitting to work for a change in average lifetimes. So with our omissions and conceivable acts in connection with health, education, homelessness, and political oppression.

The side-effects of a possible act, then, may be roughly the same as, but are likely to be different from, the side-effects of the omission. It is to be added, as in the case of main effects and experiential effects, that we are likely to have a situation of probabilities and of course different probabilities.

(xi) An agent who did act, rather than omit partly intentionally or unintentionally, would almost certainly put himself in a position different from the positions of others, partly in terms of experiential effects, partly in other ways. The others in question may be members of large classes, such as the class of adult and responsible members of a society, or smaller and better-defined classes, such as that of doctors. Omission of the unintentional and partly intentional kinds is the normal thing. The agent who acts thereby makes himself different, in particular in so far as experiential effects are concerned. He may, for example, take upon himself an unequal amount of guilt, however rational or irrational, if guilt is ever rational. The situation is quite different, incidentally, with role-omissions. There, not-acting is abnormal and unexpected, rather than acting, and it is omitters rather than doers who stand out.

Our eleventh comparison, then, comes to this: those who act, as against those who omit, almost certainly make themselves different or unequal in some class or classes of persons. The examples illustrate this.


Summary of comparisons

(i) Act and omission are both actions

(ii) Omission likely to be described negatively, act affirmatively

(iii) Omission as causally efficacious as act, although other persons likely to contribute to main effect of omission

(iv) Act fully intentional, omission any of fully intentional, partly intentional or unintentional

(v) Intentions in many acts appalling; intentions in related partly intentional omissions tolerable

(vi) Most acts and fully intentional omissions customarily accepted as somehow wrong; not so the related partly intentional and unintentional omissions

(vii) Main effects of act and omission likely different

(viii) Probabilities of main effects likely different

(ix) Experiential effects likely different, and of different probabilities

(x) Side-effects likely different, and of different probabilities

(xi) Agent’s situation likely unequal to that of others after act as against omission



We can now characterize an act and an omission of the kind with which we are concerned as (a) a pair of possible actions (b) taken as identical or similar in main effect, or at any rate connected, and such that (c) the act is fully intentional while the omission can be fully intentional, partly intentional, or unintentional, and (d) the intentions in the two actions may be greatly different in nature. We might add something (e) which has been assumed and which will be of a bit of importance later, that the omission (which we are taking to be the omitting action, or what was done) must not involve less expenditure of energy or whatever than the omitted action, what was not done, would have. The omitting action of reading the newspaper at breakfast cannot count as the writing of a whole book.

Our first issue is whether every act and omission are factually different in a single way which makes for rightness. It seems that we can now settle that issue fairly quickly. Our inquiry, a pretty exhaustive one, has turned up no single difference of fact which holds between every act and omission without exception. 

The truth seems to be that when people consider two actions with related main effects, and distinguish them as act and omission, they sometimes do so on the basis of some criteria and sometimes on the basis of other criteria. There is no single general criterion. This is not all that unusual in conceptual and indeed practical life. We might describe the situation differently and with the appearance of greater neatness, of course. That is, we might choose to specify several different senses or sorts of act-and-omission, each tied to certain criteria only. There is not much to be gained by doing so.

Our inquiry has of course produced important differences holding between some or many acts and their related omissions, but, before proceeding to them, can we conclude more generally that there is no general difference of fact which gives rise to a general difference in rightness? Can we conclude that we have not missed anything?

The first of three things to be said about this is that no supporter of the idea that there is some general factual difference, one of moral relevance, has ever made clear what that difference is. This is telling. We are not examining a matter which is essentially obscure. What we are examining, once we have got rid of such light ideas as that we are dealing with an act on the one hand and a puzzling nothing on the other, or anyway an act and something greatly more indeterminate — what we are examining is two plainish things: two actions or action-sequences. 

Certainly they may raise great problems having to do with their main effects, experiential effects, side-effects and equality-effects, and they may be baffling or mysterious in these respects. Still, we know what sort of thing is in question in these respects. If there is another sort of thing on hand, why is it not produced? As David Hume may have remarked of philosophical circumstances like this one, a challenge unmet may come close enough to a proof.

A second consideration is one which should not escape the attention of anyone who thinks of a decently wide range of examples of act and omission. We are looking for a general difference between acts and omissions which supports a difference in rightness. One can look for such a thing with more hope if one supposes, very forgetfully, that there is at least moral uniformity, as we can call it, about acts and omissions: for example, at least a presumption of some kind that all omissions are better, somehow, than their acts. 

This is a non-starter. It sometimes seems to be half-forgotten that there are more cases of act and omission than problematic ones. The fact of the matter is that some conceivable acts are unquestionably better than some omissions. Whatever one’s reluctance it is unquestionably better to act, where the act is the killing of one man oneself, if it really is true that the only alternative, one’s not shooting, will have the certain effect that someone else will shoot twenty men. It follows that while we are looking for a general difference between acts and omissions, it is not one that is always morally conclusive, in the way of making omissions better. We are looking for a general difference which is of an uncertain moral effect. Such a thing is conceivable, but not much more.

There is one final consideration which I shall postpone for a bit, until after we have considered the non-general differences between acts and omissions. With the aid of that final consideration we shall conclude that there is in fact no entirely general difference between acts and omissions which gives rise to a difference in rightness.

Let us now consider the limited factual differences between acts and omissions which we came upon in our inquiry. Some of these give rise to differences in rightness, but which are they?

The first limited difference (ii) was that there is a tendency for omissions to be described negatively and acts to be described affirmatively. It is unthinkable that this linguistic difference in itself is such as to make an action right or wrong, or more right or more wrong. The presence of a ‘not’ or something like it in a description cannot itself be of moral relevance. One may feel that the linguistic fact is somehow relevant to the matter in hand. This comes about, presumably, not only because of its connection with role-omissions, but because it is connected with non-linguistic facts, to which we shall come, which are of relevance to rightness.

The next limited difference we came upon (iii) was basically that while each of an omission and an act is a necessary condition of its respective effect, it is usually the case that the omissions of other people are also necessary conditions of the omission’s effect. Not only I but also others must fail to contribute to Oxfam in order for some persons to die. It cannot be, however, that my action can be made right, let us say, by the existence of others like it. There might be excuse for me in the fact of numbers, but it can hardly be that what I do, my action, is made right or better by it. By way of analogy, torturing would not be better if there were more torturers. 

Nor, by the way, does there seem any force in another mentioned aspect of connections between the actions and their results. If an action of mine could have prevented great distress, and did not, it cannot be that its status is improved by the fact that the distress would have occurred had I not existed.

What of (iv) the fact that the acts would be fully intentional and the omissions may be fully intentional or partly intentional or unintentional? Let us have in mind the particular comparison between acts and partly intentional omissions. These omissions are plainly of most relevance to us.

There are those who think, confusedly to my mind, that an act’s being fully intentional and an omission’s being partly intentional is of importance to the rightness or wrongness of the act and the omission. The kind or degree of intentionality makes the omission not wrong, or less wrong. There is a long tradition of commitment to such views and related ones, its greatest figure being Immanuel Kant. No doubt the tradition has things of value in it, but its central ideas are certainly not ours. That much can be shown, I think, simply by making some distinctions.

Our concern in all of this inquiry, as I remarked at the start, is the rightness and wrongness of actions. Our concern, at any rate fundamentally, is with kinds of conduct, not kinds of people. What is it for an action to be right? According to a widely-shared if not universal moral view, it is for it to be the action, of those which are possible, which can rationally be expected to give rise to the best state of affairs. Again, it is the action which would be selected by a knowing judge as likely to give rise to the best state of affairs. More might be said of such a man, but there is no possibility of giving a determinate account of his state of knowledge or his capability ofjudgement. The idea, roughly, is that his knowledge is as great as is humanly available at the time, and his judgement is the best available. In certain cases little knowledge will be available and the best judgement must be tentative or even hazardous.

As for the matter of good states of affairs, they can be as well or better described as states of experience, states involving amounts and distributions of experience. They do not have to do with any ‘values’ which are not such facts of experience, the experience of the agent and also others affected at the time and thereafter by his action. It may be wondered what ‘values’ are left out. If there is a fact about an action which in itself is not a fact of distress or frustration or enjoyment or enrichment, or any other kind of quality of experience, such a fact is not in itself of relevance to the rightness of actions. The bare facts, taken as without efficts in experience, that an action is according to a rule, or that it is thought of in a certain way by the agent, are not of relevance.

What we have, to return to the main point, is that an action’s rightness or wrongness is independent of the kind or degree of intentionality from which it issued. That an action was partly intentional, or fully intentional, or indeed unintentional, does not matter. It makes no difference to its rightness or wrongness. It is not that an omission as compared with an act is right or less wrong because the omission is partly intentional while the act would be fully intentional. 

There is a bit more to be said, however, and it will be best to proceed by considering at the same time our next limited difference (v). It was that the intentions in many conceivable acts would be appalling, while those in the related partly intentional omissions are something like tolerable. The intentions in the case of the conceivable acts in our examples would be horrific, inhuman, vicious, grasping, or something of the sort. The intentions in the case of the related partly intentional omissions are not of these kinds. On the contrary, the immediate intentions may be good ones, such as the intention to safeguard the future of one’s own children. The earlier intentions, whatever else is to be said of them, are greatly different in nature from the intentions in the case of the conceivable acts. They involve a lesser grasp of the effects of the relevant omissions.

According to a widely-shared and fundamental moral conviction, none of this matters either to the question of rightness. As remarked at the start, an action can be right, one action which ought to have been done, although it was done from the wrong motive, and an action can be wrong, something that ought not to have been done, although it was done out of a good intention. 

Consider a member of the House of Commons who voted for the establishment of the Natibnal Health Service. We suppose that he did the right thing, that his action was right, even if he was in fact opposed to the idea on principle and voted for the Health Service only out of a completely self-serving intention to gain re-election. Or rather, as needs to be allowed, there are two propositions that can be expressed by saying that he did the right thing. The main one amounts to the proposition that his action was right, understood on the lines already suggested. It is true. The other is to the effect that he was true to his principles, and so on. It is false. If we are in fact concerned with his action’s being right, however we may express this, his selfserving intention and his going against his principles is irrelevant.

It would be mistaken, of course, to suggest that the subject of agents and intentions is unimportant. On the contrary, if it is of secondary importance, it is nonetheless of great importance. It is of great importance to us since we want to have such people as will do the right thing. We are with every reason interested in the goodness or badness of people, as indicated by kind and nature of intentionality, since we are concerned with their future actions. The goodness, humanity, etc of a person is our best guide to his future conduct. 

This, however, should not be confused with the idea that what makes an action right is even partly its actual intentionality, the kind and nature of intentionality actually involved. The fact that we have a great concern with how intentional a man’s action was and what his intention was, because of our concern about future actions, should not be confused with the mistake that what makes an action right is the intentionality behind it. That question, the rightness of the action, is not touched by the agent’s intentions having been what they were, although the action will no doubt indicate at least something about his intentions, and his intentions will tell us something about his likely future behaviour. It is irrelevant to the question of the rightness of the MP’s action in voting for the National Health Service that his action, taken with a know!edge of his principles, tells us something of his actual intentions and that those tell us something of his likely future activity.

It may be wondered for a moment if I am really describing the action that turned out best, and misappropriating the term ‘right action’ for it, while the right action is in fact the action done from a good intention and so on. Some will say that the correct description of the situation is that the MP did the wrong thing, or acted wrongly, and, as it happened, this turned out well. There is nothing unusual in that, they will say.

I am not in fact confusing the right action with what turned out best. The right action, to repeat, is the one which at the time would rationally have been judged as the one giving rise to the best state of affairs. It is obvious, then, that the right action is not necessarily the action that turns out best in fact, even in the short run. Nor is there any reason to think that our common notion of the right action can somehow be reduced to that. Even knowledgeable judges make mistakes.

What has been said in definition of right actions does bring into view another indubitable connection between the conception of a right action and that of intention. It too needs to be seen clearly for what it is and for what it is not. Given what has been said, the very definition of a right action brings in a notion of intention. That is, to rewrite a bit, the right action is the one which would be favoured by a knowing judge with a certain intention: the intention to secure the best state of affairs. There is no doubt about that. 

There remains the different and essential point that it does not follow that a man’s act is right, or wrong, if he acted out of this or that intention. The MP’s act was right in that it would have been favoured by a knowing judge with a certain intention. It does not follow, as we know, that it was wrong because the MP did not have an identical intention. If, on the contrary, the MP had had an identical intention, and acted as he did, that fact of his intention would not have made his act right either, as in fact it was.

One final remark. It may be allowed that a clear enough conception has been suggested, and that we can indeed consider the question of whether actions are right or wrong in the given way. But it may be wondered if this question is the fundamental question in morality, or indeed in life. The reply must be that there is no forcing anyone to be most concerned with what is right, with what ought to happen. Most of us are primarily so concerned, rather than with the question of an agent’s actual intention taken by itself, or his moral standing taken by itself, or any other moral question. That is as much as can be said, and of course it is enough.

We were considering two questions. One was the question of whether our actual partly intentional omissions can be regarded as better than certain conceivable wrong acts because those acts would be fully intentional. The answer is no. The kind of degree of intentionality of an action makes no difference to its rightness or wrongness. Nor, to turn to the second question, does the nature of intentions matter. Whatever is to be said of the agent, an act is not made right by a good intention or wrong by a bad one. 

If it appears that these conclusions can be overturned — moral propositions though they are (4) -- this will surely be no more than appearance, since it will be owed to giving a different definition or sense to ‘the right action’, ‘what ought to be done’ and so on. This is possible, certainly, but the resulting subject-matter or concern is not ours, and not the fundamental subject-matter or concern in moral and political reflection, whether carried on in books or in the rest of life.

The next difference between some acts and omissions (vi) has to do with accepted or ordinary morality. There, there is a distinction between the conceivable acts and fully intentional omissions on the one hand, and, on the other hand, the partly intentional and the unintentional omissions. The former are wrong, the latter not so. In particular, our ordinary failure to contribute to famine relief, or to political movements against dictatorships, is not wrong, but our related conceivable acts would be wrong. 

One point here is that it would just beg the question with which we are concerned if ordinary morality were taken without reflection to be correct. Certainly we are in no way forced to follow ordinary morality. It in no way follows from something’s being generally accepted at a time that it is right. It does not follow from there being an accepted prohibition of the acts but not of the omissions that the omissions are better than the acts. What one needs to do, if one is doubtful about this proposition, is to think of the history of morality. It is a history of what we cannot take to be other than mistakes, a history of beliefs that were both mistaken and generally accepted.

We can allow, perhaps, that something or other is to be said for a generally accepted view, but it is certainly obscure what it is. It is certain that we shall not come to have a very important consideration. The obscure recommendation of ordinary morality is not enough to put into question our conclusion that ordinary morality is mistaken in the matter of the connection between intentionality and the worth of omissions and acts. 

We can also allow a second thing. It is that if people do have certain moral attitudes, however right or wrong those attitudes are, denials of them may result in hurts and distresses. If I believe that you ought to do something for me, the fact that I am mistaken does not stop my being hurt or distressed by your not doing the thing. This sort of thing is sometimes of some importance. It is not of much importance here.

Things are different with all the remaining comparisons on our list, all of them involving.limited differences between acts and omissions. All of these comparisons are of clear importance to the rightness or wrongness of actions.

It is inescapable (vii) that the harms and benefits that may result from acts and omissions are of importance to the moral assessment of those actions. This follows directly from what has been said about rightness. If considerations of death, distress, and inequality do not enter into the judgement of the rightness or wrongness of acts and omissions, then nothing does. 

Some have tried to avoid the conclusion which follows, that differences in harms and benefits must therefore give rise to differences in rightness and wrongness between acts and omissions. They have tried to avoid the conclusion that it is more wrong to have two people suffering to a given extent, or indeed a thousand, than one person suffering to just that extent. Obviously there is no avoiding it. Great tragedies are more terrible than small, and still more to be avoided. We can say, in connection with this seventh comparison, that when other things are equal we are justified in omission rather than act if the omission involves the lesser harm or the greater benefit, and vice versa.

The eighth comparison is a relevant one. An act carrying a high probability of harm, other things being equal, is worse than an omission carrying a lower probability of an identical harm, and vice versa. There are a number of related propositions about probability. An omission carrying a certain low probability of a certain greater harm may be worse than an act carrying a certain high probability of a certain lesser harm, and so on.

The ninth fact of acts and omissions, their experiential effects, is also morally relevant. In certain instances these effects may be of decisive moral importance. In other instances, perhaps more, they are less important than other things. Side-effects, the subject of our tenth comparison, are also morally relevant, and perhaps likely to be more important than experiential effects. Finally, it is clear enough that the eleventh comparison, having to do with the position of the agent in some class of comparison, is of relevance. There are some actions which a man might contemplate, and which it would be wrong for him to do because of the extent to which he would be made unequal as a result.


We now have a grasp of the comparisons between acts and omissions. Let us turn back to the particular omissions and conceivable acts with which we began. Would the acts be worse than the omissions are? Would they be no worse? We should now be able to decide. If they would be no worse, the violent have truth on their side in the claim that our ordinary law-abiding conduct is wrong. Their tu quoque, whatever is to be said of the relevance of such retorts, is not to be put aside as a falsehood or a mistake. Let us get the answer to our question.

We have spoken of ‘the omission of not sending what money we could to Oxfam’ and the conceivable act of ‘sending poisoned food parcels to India’. There was also the omission in which we ‘fail to contribute funds to political movements against dictatorships’, and the conceivable act in which we ‘contribute funds to a dictatorship’. These easy phrasings may lead one to overlook what has not been overlooked in our survey, that omissions and acts can differ in main effects. 

They can differ, as already remarked, even when two different main effects fall under the same piece of language, the same description. Certainly that is clear enough when one stops to think. On the other hand it is possible that omissions and acts be identical in main effects. There is precisely the same situation with the other important comparisons (viii to xi) between acts and omissions. There can be difference or sameness with respect to probability of main effect, and with respect to experiential effects, sideeffects, equality-effects and the probabilities of each of these.

There is another fact, a connected one, which is fundamental. We can choose what conceivable actions to compare with our actual omissions. Given an actual omission, the related conceivable act is not thereby determined. There is not one conceivable action which is the act for that omission. In fact there is no fixed answer, or rather, there is no answer, to this question: What is the conceivable act which stands to a particular omission? Habit, customary examples, and simple situations may lead us to suppose that there is an answer, but they mislead us. We choose which possible act goes with an omission, what to compare with what. We can vary examples in order to find things out. We have a freedom of inquiry.

Perhaps the tendency of my argument will now be anticipated. We tend to think that conceivable acts would be worse than our omissions because we have in mind certain of those conceivable acts rather than others. If we think of others, there is the plain conclusion that our omissions are as wrong as those acts. Those acts would be terribly wrong.

Consider once again the first of the six examples as it was given. In thinking of it, we are very likely indeed to have in mind several ideas. One is that sending poisoned food would not merely probably but almost certainly have a terrible main effect, deaths that otherwise would not happen, while not contributing to Oxfam would probably have some terrible main effect, some number of deaths that otherwise would not happen, not necessarily the same number of like deaths. Here there is a difference in main effects, and also in their probabilities. As for the remaining three comparisons, we are likely to think of worse experiential effects for the agent in the case of the conceivable act, worse side-effects, and a worse equality-situation. There can then be no surprise in the fact that we have a strong inclination to think of the conceivable act as worse than our actual omission.

There is another comparison. Let us fix in our minds a conservative estimate of the whole upshot of an actual omission and then conceive of an act of the same upshot. How much could I contribute to Oxfam next year? Let us say 4,000. Suppose I do not give the money. 

It is a secure judgement, surely, that my not doing so will have at least two deaths as its very probable main effect. Let us say, for purposes of argument, that the probability is 75 per cent. The experiential effects on me of not doing so will be, very probably, small. My own equality-situation, vis-à-vis those of people like me in my society, will be pretty well unaffected by my omission. The side-effects of my omission will be considerable: they will include the effects of what I do instead with the 4,000, and also the effects of the two deaths. Suppose, for purposes of argument, that the money will be spent on improving the life of myself and my family. These improvements, by way of certain comparisons between societies, will provide what indubitably must count as indulgences and luxuries for myself and my family. The effects on people related to the two who die will be very different of course.

It is conceivable that I arrange to perform a certain act, perhaps a poisoning, carrying a 75 per cent probability of the deaths of two persons. If I were to do so freely, the experiential effects on me might very probably be small, not different from those of my actual omission. Let us have precisely this possibility as to experiential effects in mind. I would not then be the person I now am, of course, subject to certain feelings of guilt and so on. Let us suppose also that the side-effects by way of the two deaths would be the same as before. That is, we suppose it is by my act that I and my family have 4,000 worth of indulgences, and that there are the same side-effects as before by way of the deaths of the victims. Let us suppose, finally, for reasons which we can conceive, that my own equality-situation in my life would be the same as in the case of my actual omission. Certainly we have now conceived of an entirely extraordinary situation, one that is certain never to obtain. That does not matter at all for the argument. What matters is the likeness of this merely conceivable act to my actual omission, and what we say about the conceivable act.

What we must say, of course, is that my conceivable act would be wrong, and thus we come to the conclusion that my omission next year will be precisely as wrong. We come to the general conclusion that our ordinary lives consist in omissions as wrong as certain conceivable terrible acts. Only quite different conduct, of a kind sometimes mistakenly put aside as something for saints, would be right. (5) 

It is to be allowed, of course, that my spending the 4,000 on myself and my family will only be as wrong as the one conceivable act. It is not that it will be as wrong as some multitude of conceivable acts each similar in upshot to one of the multitude of omissions I will perform in spending the 4,000 as I shall. It does not follow from the fact that my doing A was also not doing a thousand other things, that I could have done the thousand things. That does not much reduce the gravity of our general conclusion.

Is it reduced, incidentally, by the objection that my omission will be only about half as wrong as the conceivable act, since the conceivable act would also be something else: an omission to contribute? (6) My spending 4,000 on myself and my family, it may be objected, is only harmful as an omission, but the act of poisoning would be harmful both as that act and also as an omission, the omission of not contributing money to Oxfam. The act of poisoning would be about twice as wrong.

There are possible situations in which an objection of this kind is in place, not that its conclusion is greatly reassuring. (The omission in those situations, it might be replied to the objector, is fully half as wrong as some terrible act.) The objection is in fact not relevant to our argument. 

This is so for the reason that the act of poisoning would not necessarily count as a relevant omission. It would not count, for example, as omitting to contribute 4,000 to Oxfam. This follows from our conception of acts and omissions. An action is a given omission only if what was not done could have been done with no greater expenditure of energy, resources and so on than that expended in what was done. My reading a newspaper for a few moments cannot count as omitting to write a book. My conceivable act of poisoning, to return to that, would not necessarily be anything at all like omitting to contribute 4,000 to Oxfam.


The conclusion that our ordinary omissions are terribly wrong depends on there being no differences between acts and omissions other than those already listed. It depends on its being true, more particularly, that there is no general factual difference between acts and omissions. Two arguments were given against there being such a thing, and a third promised. It can now be provided.

If the factual difference exists, then even if it escapes our detection or description, we ought to be able to feel its moral effect. We ought to be aware of its effect in terms of the rightness or wrongness of actions. Such responses could not possibly be a matter of the special capabilities of only certain persons. Morality is not ESP. The difficulty, of course, is that we have to be sure, about any example, that we are not in fact responding to relevant differences of the kinds we now know, such as difference in probability of main effect.

We can devise certain examples, in areas less controversial than those we have been considering, and hence where we are less likely to go wrong, and with these examples it is impossible to have any feeling of difference in rightness between act and omission. The examples, of course, will be such that there is no difference of the kinds so far allowed to be relevant. We can proceed by way of a euthanasia example. There is the possible act, which might be a doctor’s turning off a tap in a blood transfusion tube. The patient, who is incurable and in torment, dies. There is the possible omission, which might again be not providing penicillin for pneumonia. Again the patient would die. 

If we hold the act and omission identical in the relevant ways of which we know, it will surely be impossible to achieve any sense of difference in rightness between the two actions. Certainly it is not hard to drift into the feeling that there is more to be said for the omission. It seems invariably true that if this is done at all reflectively, it is done by changing the example: that is, by not holding the two actions identical in the five relevant ways of which we know. The temptation, if that is what it is, seems near to irresistible. In my view all recent philosophers who have tried to claim a special moral difference between acts and omissions have fallen prey to the temptation, or anyway not taken good care to have their examples satisfy the condition of having act and omission alike in the given ways. If the two actions are really held identical in the given ways, I propose, we cannot achieve any real sense of moral difference between them.

Some will suppose there is another way of resisting the general conclusion that in our ordinary omissions we do wrong to a terrible extent. They will suppose that the conclusion can be refuted or put into question by this argument: if the conclusion were true, we would be moral monsters; the wrongfulness of our actions would surely support this conclusion about ourselves; we are not moral monsters; indeed, it would not make sense to claim this of all of us; therefore the conclusion that we all do wrong to a terrible extent is mistaken or somehow in question.

The issues raised are large ones, and I shall do no more than quickly suggest reasons for thinking that the objection fails.

It has already been remarked that it does not follow directly from an action’s being right that it preserves the agent’s moral standing. Nor does it follow directly from an action’s being wrong that it reduces or destroys the agent’s moral standing. Still, from an act’s being wrong together with certain other facts it does follow that the agent is open to question, or worse than that. What can be said in the case of our omissions?

It is to be kept in mind, first, that few if any of us omit to contribute to Oxfam in anything remotely like the fully intentional way. We do not aim at causing death. There are no such monsters. Secondly, are there only a few people, as may be thought, whose omissions are of the opposite kind, wholly unintentional? Are there only a few, that is, who have not earlier taken avoiding action? There are more of these innocents than might be supposed, at least if one gives a certain characterization of what is required for partly intentional omission.

It is reasonable to say that for something to count as an earlier intention-to-avoid, the person in question needs to have had more than fragmentary ideas in mind for a moment, more than a flicker of images. The person in question needs to have had a decent awareness of the relevant effects of the later omission. Also, he needs to have had a decent awareness of what he might do instead in the future. There are many people who have too small a conception of what might be called their world of possible effectiveness. They do not suppose, for example, that they can do anything to contribute to less inequality in average lifetimes. Nor can they be censured for their weak grasp of their possibilities of action. There are therefore more of the innocent, the unintentional, than might be supposed.

Consider now the moral standing of the partly intentional. Putting something out of mind in the required sense is likely to require self-deception. This cannot consist, as some philosophers seem to believe, in successfully lying to oneself, and therefore in believing both p and not-p together. Self-deception, at any rate when it is non-pathological, consists rather in an avoidance of evidence, or of pointers or clues, with the aim of avoiding belief. It Consists in persisting with a question, which is better than getting a particular answer. Part of what enters into putting something out of mind, then, is the fact that we have kept a good distance from places of evidence, perhaps evidence about average lifetimes in developed societies, or the state of education in certain sorts of school. (I have not said enough, of course, to give the reality of all this. There are many ways and styles of avoiding evidence.) 

To come to the essential claim, there is some exculpation possible for typical self-deceivers. If I have the belief that a terrible state of affairs exists, about which I could do something, that is different from my in fact having a question about such a state of affairs.

There is something else to be said of partly intentional omission. Omitting to contribute to Oxfam, to the extent that one is in a state of belief, is likely to involve beliefs as to distant awfulness. It cannot be true that an act or omission is right or wrong because of its spatial relation to its result. It cannot be that our moral obligations are in any way a function, simply, of mileage. But the questions of acts is not the question of agents, and the question of agents is not easy. 

What about the relevance of distance, in the ordinary sense and also others, to our standing as moral agents or as human? It seems to be true that those who drop bombs are less morally revolting, if those are the right words, than those who do the close work of hand-to-hand killing. Again, one would feel differently about a family, say, who let a man starve to death in their own flat, as against a family who failed to send food to a man known by them to be dying in another country. What I want to suggest, then, is that the fact that the victims of omissions are distant does have some exculpating effect with respect to agents.

Further and finally, with respect to partly intentional omission, it is safe to say that in ordinary morality it is as if there were some great difference of fact between acts and partly intentional omissions. This is a common belief or conviction. In ordinary morality, further, there is no awareness of the great wrongfulness of our ordinary omissions. These are both mistakes, I think, but they are not mistakes to which moral blame attaches. It is not a general requirement of being a good person, now, that one does not make these mistakes.

I take it that these various considerations go some considerable way towards exculpating many people. Or, to come to safer ground, and all the ground I need, it seems that the considerations do the lesser thing of absolving most people from being moral monsters. Whether exculpation, or the lesser absolving, it is partly something that is true at this time in human history. It is dependent on a certain general awareness at this time. It is not an exculpation or absolving that will persist, as it is, into the future. One may wonder, of course, whether it may apply now to oneself, but that is something else.

We were considering the argument that if we in our omissions do wrong to a terrible extent, we are moral monsters; and we are not such, and so it cannot be that we do wrong to a terrible extent. The reply has been that for various reasons a general and substantial proposition about overwhelming guilt on the part of many people does not follow from, and hence cannot endanger, the conclusion about the great wrongfulness of our omissions. It is not entirely clear that the conclusion would in fact be endangered if it did carry the given large consequence, but I leave the matter there.


By our ordinary omissions we do as wrong as we might by certain awful acts. Some of us or indeed most of us will pay this conclusion no serious attention. Perhaps it is true that no one will take it so seriously as to act upon it. Such results are dispiriting only to those who are unaware of the history of political philosophy, of serious moral protest, and of like things. Virtually every outlook or institution or practice which is now venerated did in fact begin life as a silly proposition upon which no one acted. It began life as such a proposition although in fact there was enough argument for it, and not enough against it.

What we must accept, to look back to the beginning of our inquiry, is that those who engage in violence of the Left have a truth on their side when they respond to the charge that they do wrong with the retort that we do wrong. What is to be said of this tu quoque? I supposed in the beginning that it cannot be put aside, and promised to return to the matter.

Overstatement is possible. Some of the violent will say that in fact they are not limited to the tu quoque, whatever it comes to. There is more. They will say that our ordinary contribution to the frightful circumstances of shorter lifetimes, miseries, and the want of freedoms does not issue only in the judgement that we do wrong, and whatever follows from that, but also in the judgement that our ordinary condemnation of violence is incoherent. If this is not neatly expressed by those who defend their violence, that is in a way not unusual, and in a way not essential. It is not unusual for those who act in certain causes to be unable to do well in arguing for those causes; their being unable to do so is not essential to those causes being defensible ones. Policemen and soldiers should come to mind.

The claim by the violent that our condemnation of violence is incoherent proceeds as follows. We judge our kind of conduct to be right, and the conduct of the violent to be wrong, but the two kinds of conduct are alike in a certain fact, and this fact of likeness is allimportant. Each kind of conduct consists in the denial of human needs and indeed of life. On the one hand these things are denied to the primary victims of acts of violence, and also in part to secondary victims, principally the families of primary victims. Whatever else is added, this much must be allowed. On the other hand, as has just been said, there are the innumerable victims of those of us who live ordinary law-abiding lives. They live less long, and badly, and without full freedom. 

On each hand, then, there is the denial of human needs and life. It is also true that while the two kinds of conduct are alike in the given fact, there is no relevant way in which our ordinary conduct is different from that of the violent, no difference which makes our conduct right and that of the violent wrong. We know there is no such difference in the fact that our ordinary conduct consists of omissions and the conduct of the violent consists of acts. There is the result that our ordinary condemnation of violence is incoherent, that we fail to make what can count as judgements at all. Our utterances are as insubstantial as those of a man who ‘describes’ an object in a certain way, perhaps as being of a certain colour, simply and solely because it has a quality to which he points, and then gives a contradictory description of another object with precisely the same quality. He has, in the most important sense of the words, said nothing at all.

The argument does not work, and in fact it overlooks plain things established in our reflections on acts and omissions. We have found that there are conceivable acts which are factually like our own omissions in all the relevant ways. Our omissions, therefore, are on a level of wrongness with these acts. None of that, however, goes any way towards establishing that our omissions are in the relevant ways like the acts of the violent. That is another matter. There is in fact no likelihood that our missions can be shown to be like acts of the violent in main effect, or probability of it, or side-effects and their probability, and so on. The claim of incoherence can in fact be defeated by pointing out that it is just mistaken to say that our ordinary conduct is in no relevant ways different from that of the violent. It is different in ways in which omissions may differ from acts.

The violent, then, have only their tu quoque, and not the claim of incoherence. But is there something, perhaps much, to be said for the tu quoque?

The violent, when they are taxed with what will be called the irrelevance of their claim that our ordinary conduct is wrong, can maintain the strong truth that our two kinds of conduct are connected. It is not that the terrible circumstances of misery and injustice, against which violence is directed, come about by chance. They have not come about through historical passages in which we and our predecessors have played no particular part. The circumstances of misery and injustice, rather, are as good as our own work. We contribute to them essentially by our wrongful conduct. It is not that there is no connection between violence and our omissions, but rather that there is the connection that our omissions are essential contributions to the misery and injustice against which violence of the Left is directed.

More can be said along these lines, but all of it may get the reply that it is nothing to the point. Our ordinary omissions, it will be said, are not the subject-matter. The subject-matter is violence and what is to be said for and against it. It may be that there is the claimed connection between our omissions and the violence, and, for purposes of argument, it can be granted that our conduct is wrong. It is none the less true that the subject-matter is what it is, and that it is not our omissions.

There is only one proper judgement on this dispute. It is that there is no ordained subject-matter. There is never an ordained subject-matter. At bottom the criterion of relevance for any discussion is fixed by the demands of those who are taking part in it. If they cannot agree, they cannot discuss. No doubt anything which is one discussion remains within some wide boundary. No doubt it is true, too, that fact and logic impose requirements of appositeness on answers to particular questions, replies to particular claims. There is no fact of the world, however, and no consideration of logic, which stands in the way of the demand of the violent to bring our omissions into the discussion. It is not an absurd demand. What can stand in the way of it is only an opposed demand.

My own determination will by now be clear enough. It is that the discussion be the wider one. 

My first reason is that it seems possible that we proceed in too simple a way if we declare that the wrongfulness of violence is in no way diminished by the wrongfulness of our omissions. No doubt it is worth someone’s while to say that two wrongs do not make a right. It may also be worth saying that there is reason to be sceptical, or perhaps it would be better to say that there is reason for some reserve, in a certain kind of situation. It is the kind of situation where we have one man’s judgement on another’s conduct, and the first man may be influenced by an unhappy perception of his own relation to the other’s conduct. More particularly, we have some awareness that we make our necessary contributions to the circumstances which have led and continue to lead to violence. It seems to me a likelihood that this awareness affects our condemnations. Our condemnations, although condemnations still, are likely to be different if we are more aware of our own conduct and its wrongfulness.

There are two more reasons for the wider discussion. One is that it is in fact more important that we change our ways than it is that the violent change theirs. We do greatly more damage. Another lesser but good reason for the wider discussion is that it may do a little to diminish violence. The narrower discussion, rightly, has less chance.


1. See the admirable book by Jonathan Glover, Causing Death and Saving Lives (Harmondsworth: Penguin, 1977) particularly Ch. 7, and an alarming short article, ‘The survival lottery’ (Philosophy, 1975) by John Harris.

2. The example is owed to Philippa Foot, who writes that ‘most of us allow people to die of starvation in India and Africa, and there is surely something wrong with us that we do; it would be nonsense, however, to pretend that it is only in law that we make a distinction between allowing people in the underdeveloped countries to die of starvation and sending them poisoned food. There is worked into our moral system a distinction between what we owe people in the form of aid and what we owe them in the way of non-interference’ (‘The problem of abortion and the doctrine of double effect’, The Oxford Review, 1967).

3. In Utilitarianism: For and Against (Cambridge: Cambridge University Press, 1973) Bernard Williams discusses the case of a man who must choose between the act of shooting one political protester himself, or omitting to do so, with the result that someone else will shoot twenty.

4. See my A Theory of Determinism: The Mind, Neuroscience, and Life-Hopes (Oxford, 1988), pp. 420 ff.

5. In ordinary morality there is a distinction between what we are obliged to do, and what is done by those who are or used to be called saints and heroes. They are those individuals, such as Albert Schweitzer, who are said to be moved by an ideal, and so to do more than their duty. It is said that they do things of moral excellence, but they would not do wrong if they did not do them. They do some of the things that we omit to do.

All this is mistaken if the argument I have been supporting is correct. Those who are commended as going far beyond their duty may not be doing much more than just their duty. They are doing things such that if they did not do them, they could be reproached.

Ordinary morality, with the distinction just noticed, has not been without its advocates. Mr. J. 0. Urmson is one. He tells us (‘Saints and Heroes’, in A. L. Melden (ed.) Essays in Moral Philosophy, Seattle, 1958) that ‘it would be quite ridiculous for everyone, however circumstanced, to be expected to go off and nurse lepers’. Does he tell us, however, why a somewhat more sane expectation, but of the same impulse, would be wrong?

He does declare that it is the duties of ordinary morality, which of course allow omissions of the kind we have been considering, that are the important thing: they secure that life is not ‘brutish and short’.

The phrase is unfortunate. ‘Brutish and short’ lives, for some, are guaranteed precisely by our not going beyond duty, so-called, by our persisting in some ordinary omissions.

A second of his contentions is that duties more stringent, more in line with the argument I have been supporting, could not be kept by us. All duties, including the less stringent, would then come to lack urgency. We would come to have that supposed upshot upon which scores of Utilitarian arguments rest, and whose improbability undermines them, ‘a general breakdown of compliance with the moral code’.

The other relevant contentions, in my view, are about as hopeless. The third is that a moral code be made up of rules whose complexity is not so great as to be unmanageable; the fourth is that we cannot have a morality which involves the possibility of too much expectation and too great demand on individuals; and the fifth, not entirely unrelated, is that free choice in moral matters, rather than the constraining weight of obligation, is better wherever tolerable.

6. This possible objection was suggested to me by Janet Radcliffe Richards. 


This paper was published under the title 'Our Omissions and Their Violence' as a chapter of my book Violence for Equality: Inquiries in Political Philosophy. More of the morality of the paper and of Terrorism and Philosophy 1: On Inequality and Terrorism, and Differences We Make Between Them is given in four other papers that can be consulted on this website: Consequentialism, Moralities of Concern, and Selfishness; Hierarchic Democracy and the Necessity of Mass Civil Disobedience; What Equality is Not, Fortunately; and What Equality Is.