Chapter 8: The Pauline Principle

We are but imperfectly rational and imperfectly moral. Almost all moral philosophising is the attempt to close the gap between the rational (or apparentlv rational) and the moral (or apparently moral). It is not often that the attempt is entirely successful. The casuists decided that a strict interpretation applies in “afflictive” matters and a broad interpretation in “favourable” matters. That is, rules and maxims are to be interpreted narrowly (to minimise the area of application) where their application is burdensome but are to be interpreted broadly where that application has a favourable outcome. This policy is intellectually dubious but is morally creditable: the casuists saw that people were caught on a hook (in many if not most cases an unnecessary one, fashioned from the mistaken idea of “intrinsically evil” acts) and tried to get them off it at the price of a bit of intellectual manoeuvring.

The attempt to close the gap between the rational and the moral typically manifests itself as an attempt to discern the intelligibility of action as moral or immoral. The teleologist/consequentialist locates intelligibility in the consequences of action. The deontologist begins elsewhere. Neither has any touchstone for distinguishing truth from error but the touchstone of coherence: once a contradiction appears in a train of argument, error is known to have entered. But it could be a serious mistake to allow the intellectual to sit in judgement on the moral, probably more serious than the mistake of allowing the supposedly moral to dictate intellectual propositions or conclusions. Very likely the best we can do is to try to use either to correct the other while keeping a wary eye on both. Ideally, the moral would rule the intellectual, taking it that the spirit is nobler than the mind or, to put it another way, taking it that fully moral conduct represents perfect rationality. But we are not reliably moral, any more than we are reliably rational.

The principle of the double effect is another example of the morally creditable and intellectually dubious. It seeks to resolve certain difficulties that a literal interpretation of the Pauline principle would raise. Does the name convey the idea of a rather technical sort of thing, something concerned with academic distinctions of interest, perhaps, to theologians or to moral philosophers but irrelevant to the ordinary person who is simply concerned to live his life usefully or at least not too discreditably? Do not be misled. The principle concerns what to do when one must choose between courses of action all of which seem to be morally objectionable from some point of view or other. But moral choice concerns almost nothing but this kind of choice. The principle has been described by Peter Knauer (“Readings in Moral Theology No.1” Eds. Curran and McCormick) as “the fundamental principle of all morality”. Those who dismiss it as casuistical codswallop show – since they have offered nothing better, and although they may be too polite to describe it thus – that they have failed to get to grips with the problem it seeks to address. If there is anything wrong with it, the matter is a serious one. If discussion of it is somewhat tedious the tedium cannot be helped. The question is too important to be ducked (and any conclusions reached on an intellectual basis should be brought to the bar of moral judgement before being accepted).

To begin with the principle of the double effect rather than with the Pauline principle itself may look like putting the cart before the horse. Heuristically, however, it may be the simpler approach. The principle of the double effect is the Pauline principle in application and examination of it may bring any shortcoming to light more clearly than would a direct examination of the Pauline principle itself.

Richard A. McCormick indicates (“Doing Evil to Achieve Good”) that the generally – accepted formulation of the principle of the double effect is along the following lines. An action may be regarded as justified, even where it causes evil, where four conditions are satisfied:

1) The action is good or indifferent in itself, that is, it is at least not morally evil.

2) The evil effect is sincerely not intended.

3) The good and evil effects must be equally immediate from the causal point of view.

4) There must be a proportionately grave reason for permitting the evil.

It will be necessary to look at each of these conditions.

The first condition is that the action is good in itself, or is at least not morally evil. Now, what does it mean to say that an action can be good, or can be morally evil, in itself: that an action simply as such can be morally evil?

Ask first: what does “evil” mean? Ask second: what is an “action”?

“Evil” may mean (non-moral) harm such as pain, hunger, illness, death. Or it may mean “doing wrong”, in which case to say that evil may not be done is merely analytically true and is hardly worth saying. But is it always “doing wrong” to bring about harm for the sake of what is taken to be a more important benefit? Or sometimes wrong? Or what are the cases, if any, where one may seek to achieve a greater good at the cost of a lesser evil and still “do wrong”?

The sheer occurrence of harm, without human volition, may be regarded as non-moral. However, it is (presumably) uncontroversial that the deliberate visitation of harm on someone is always morally significant and hence must be either right or wrong. Again, “evil” may refer to moral evil understood simply as such (e.g. malice, the deliberate infliction of harm on an innocent person without the purpose of a commensurate benefit to him, the conscious disregard out of selfishness, etc. of another’s rights or claims). It will simplify discussion to take it that moral evil in some such latter sense may never be done.

The first condition cannot refer to evil in the sense of non-moral harm for the doing of it brings (non-moral, as such) harm already into the moral sphere and whether it is good or bad in the moral sense is something that cannot be settled merely by looking at what is done “in itself”. A Kantian should have no difficulty with this. Kant held that an action cannot be good merely in itself, or even as done in accordance with an aim of general happiness or general harmony. An action is good only as performed in accordance with perfect rationality, as willed for the sake of the moral law itself. If Kant is right it seems to follow that acts, merely as such, cannot be evil since neither can they, merely as such, be good. And “simple” moral evil (we take it, since this is again analytically true) may not be done on any condition or any set of conditions. So unless it refers to “evil” in some third sense it looks as though the first condition is one impossible to fulfil. What third meaning could there be?

For we need to be clear about it: any human act, merely as a human act, aims at achieving some good. It may be the good of serving one’s self-centred or immediate inclinations, or of serving another person, or of serving an ideal or some large and distant purpose. It may be the good of deflecting a foreseen harm from oneself to another (and innocent) person or even the good of indulging one’s inclination for malice or revenge against another person or for domination over another person. Whatever it is, as a human act, what is done aims at some good. The will simply cannot choose what is perceived to promise no benefit whatever.

So any and every human act is done that good may come of it, in an inclusive sense of “good”. This is what defines it as an “action” as against, say, snoring, or looking at the box while thinking of something else. And virtually every human act aims at good at the price of something harmful or at least disagreeable, even if only fatigue, or going to work when it would be nicer to stay in bed. The first condition cannot refer to action that is good “in itself” in the sense of action that achieves a benefit at no price whatever: few actions do, perhaps none. Nor can it refer to action that produces good for oneself at the price of harm for another; the moral tradition concerned does not so identify it and anyhow it is easy to think of circumstances (e.g. self-defence) where one would take it to be not wrong to do what brings harm to another and benefit to oneself. So?

If you didn’t know already, you might never guess. There are two assumptions on the basis of either of which the first condition is at least intelligible. One is that there are certain absolute human rights, rights that inhere in human persons as such and that in consequence may not be infringed in any circumstances without doing wrong, that may not be abdicated or alienated. Something is said elsewhere about this idea. Those who are not hung up on it can see, as soon as examples are produced, that the idea is unsustainable. Life itself is a pre-condition of the enjoyment of any other rights whatever, yet we are quite clear about it that a person may forfeit his right to life in certain circumstances and that organised society depends, in general, on rights being conditional and not absolute. One example only: the proponents of absolute human rights are unable to say, clearly, why the right to be saved of the one pursued takes precedence over the right to life of the murderous pursuer. The second assumption is that certain acts are evil in themselves, either because directly forbidden by God or because unnatural and hence indirectly forbidden by God. While this idea is rooted in a certain theology, outside of theology the following can be said: an action is not meaningfully to be referred to as a human action without a reference of some kind, implicit or otherwise, to motive or presumed motive, intention or presumed intention, purpose or presumed purpose. The idea that an action can be a human action in the proper sense without any of these is merely unintelligible. An involuntary human act, such as a sneeze, is not in that sense a human action and a voluntary act must incorporate purpose etc. Yet that idea is what is sought to be conveyed by the suggestion that a human action can be either good or evil “in itself”. The confusion does not end there. The first condition of the principle of the double effect does not, as one might now suppose, simply rule out those actions that are held to be evil in themselves and then leave the justification or otherwise of other actions to be settled in accordance with the remaining three conditions. To a significant extent it seeks, through the application of those other conditions, to legitimate certain actions that, under a simplified categorical description (e.g. killing), have already been declared to be evil in themselves. Thus it seeks to square the circle.

To take the example of killing: death, simply as death is, we may take it, a non-moral evil. However, killing someone is certainly not non-moral. It is either right or wrong. Thus self-defence cannot satisfy the first condition, that is, the action of self-defence consisting of one’s injuring or killing an attacker cannot in itself be either good or indifferent, apart from any other consideration. In the traditional thinking the supposedly intrinsically-evil acts centre on killing and on sex. From the start certain kinds of killing, as in self-defence, were held to be justified. The principle of the double effect was developed to explain how to justify an action supposedly wrong-in-itself, and argument goes on to the present day about its application in cases involving sex and reproduction and in cases involving methods of warfare.

The second condition is even trickier. The evil that is done must not be intended. It may be foreseen, but merely tolerated, merely permitted to occur, and must not be directly willed or intended. The expression “indirect voluntareity” is sometimes used.

What this means seems clear enough in one kind of case, that is, the case where the unintended evil depends on the volition of another agent or of other agents. Perhaps the best-known (and best) example is that of the sheriff and the mob. The sheriff foresees that, unless he hangs an innocent man, the mob will riot and others, perhaps numbers of others, will be killed. He does not intend that the riot should take place. He does not intend the ensuing deaths. But he refuses to hang an innocent man, and we agree with his decision. The evil that is not to be done is the deliberate killing of the innocent and the good that might ensue is the preservation of the lives of those who will be killed in the course of the riot (the point can be explained thus: questions of moral right and wrong cannot even arise except for persons who are held to be responsible for their own acts. The sheriff is responsible if he hangs the innocent man. He is not responsible for the riot – others are. What he foresees is what other persons will do but what he is responsible for is what he himself does).

But what could it be, to perform a voluntary act, and yet not to intend its foreseeable consequences where those consequences depend only on one’s own act? How can one merely “permit” the effects (or some selected effect or effects) of one’s own deliberate act? In our own day some of the difficult questions centre on abortion. Roman Catholic teaching forbids removal of the foetus (alone) in the case of an ectopic pregnancy, on the grounds that the foetus is a human person and that his removal from the Fallopian tube is murder. However, removal of the Fallopian tube itself (with foetus) is permitted in such cases on the grounds that the tube is in a pathological condition posing a threat to the mother’s life and that removal of the foetus is an indirect or incidental or unintended side-effect. Yet the latter operation mutilates the woman unnecessarily (for medical skill cannot at present save the foetus anyhow) and the argument does not address the questions: what is it that is held to justify this mutilation? And, since necessary killing in self-defence is held to be justified, why is self-defence prohibited in this case? (The innocence of the foetus is not to the point, any more than is the putative moral innocence of a homicidal maniac).

It is worse than that, indeed, for the same teaching forbids tubal ligation (because this is direct sterilisation) and yet permits hysterectomy with few enough restrictions in practice (hysterectomy being a far more damaging procedure that indirectly results in sterilisation). Yet hysterectomy results in sterility indirectly, and not directly, only on the basis that the purpose of the operation is to prevent further damage, e.g. via another pregnancy, to an already-damaged uterus – damage that could with far less harm be averted by the former procedure. It may reasonably be suspected that actual removal of reproductive organs is unnecessary in every relevant case. Where that removal is not indicated on medical grounds alone it seems inescapable that a more or less deliberate misrepresentation of intention takes place. And a tough-minded critic would have some ground for arguing that, where that occurs, what is done is simply and obviously immoral: a self-serving prudence, a playing safe by way of rule-observance, a career-oriented complaisance, an abdication of responsibility for personal choice at the cost (to another person) of unnecessary harm – any or all of these. A military commander could likewise play safe by attacking a position with a blanket of missiles or bombs (incidentally killing some non-combatants who are known or are reasonably thought to be there) rather than by a less destructive ground assault – assuming that the latter represented a reasonable military option – that would almost certainly entail the more direct killing of the non-combatants concerned. So is there any justification? Or is immorality involved?

An answer – it may or may not be the answer that the relevant authorities would make but it seems at least arguably well-grounded – is: the need to hold the line. It is an answer that brings one back to the almost (almost, though not quite and not everywhere) overwhelming importance of rule. Rules are not only enormously important: they must also be fairly simple, fairly black-and-white. Complicate a rule by admitting this exception in this particular case – and immediately it becomes understood that the rule is not itself the principle but that the principle involved is something …. well, it isn’t quite easy to say ….. and rule, principle and all may be carried away.

The point calls for some notice, even at the cost of further tedium, because it is central to the Pauline principle and to traditional morality itself. Anglo-American moral philosophy is done mainly in universities and sometimes manifests poor contact with reality. Some exponents are inclined to take a high moral line about the above Roman Catholic position. However, Roman Catholic moral theology is (or used be) done, via the confessional, in the roar of the streets and in the clamour of conflict. It is acquainted with reality to bruising-point. It has been a long haul to narrow justifiable homicide to the currently-recognized categories. The need to uphold rule, and the need also to recognise exceptions in appropriate cases, generate problems whose optimum solution can be a delicate business – and here we get somewhere near to the heart of it. Perfect solutions may be sought, and optimum solutions may be hoped for, but perfection is not reasonably to be expected. We are not yet perfected. Casuistry was careful about two considerations in particular when the question arose of recognising an exception to a rule: the exception must be clear, must not entail the risk of confused application or of illegitimate application obfuscated as legitimate. And, even more important, the features of the exception must not be elective (for then one could contrive those features, by manipulation of events / arguments / definitions, so as to apply the exception in accordance with desire rather than in obedience to law).

No one knows, or could know, at what point a fertilised human ovum becomes a human person. Any point at all that might be chosen subsequent to fertilisation can represent only a guess, almost as arbitrary as the point that may be chosen to divide the night from the dawn. “Viability” is no test: a new-born baby is no more “viable” than an unborn baby. Rationality is no test, for some human beings never attain to rationality. There is indeed a strong case for abortion where the alternative is the mother’s death, but it is not self-evidently a conclusive one. Grant that exception: then why not grant exception where the alternative is ostracism and social extinction? Where the expectant mother would apparently prefer to die? Where the alternative is financial burden? Or inconvenience? After all, just what principle is supposed to be at stake if abortion is permitted at all? Where can the line be held? The values, so far as identifiable, shade into one another. If abortion is permissible on certain grounds, why not also infanticide on the same grounds? That is, why should the biological fact of birth confer a status that the more fundamental biological fact of fertilisation does not confer? (We saw that social contract theory can answer this last question. However, we also saw that social contract is a maimed ethic).

Rational argument can support the Roman Catholic position (apart, perhaps, from the case where self-defence can be invoked) if one assumes that human personhood begins at fertilisation (and not merely human life, since there is human life in one’s little finger). There is rational ground for holding that the “negative” rules (not to murder, not to steal, not to lie etc.) bind more strictly than the positive rules (save life, help those in need etc.). Indeed, the argument is virtually incontrovertible. This implies that, given the assumption mentioned, saving the mother’s life is less important than not killing the child.

It seems probable that the proponents of this conclusion don’ t actually like it any better than does anyone else. They have at least tried to be clear-headed, on the assumption that one must first settle at what point human personhood begins. But it is not enough for those who disagree merely to disagree, nor is it enough to adduce some ground of disagreement. Any ground of disagreement must be defended all the way down, in the face of all of its implications and consequences, and no convincing ground appears yet to have been adduced. However, what sometimes seems to happen in the case of those who argue for abortion is that they simply haven’t made up their minds as to whether the foetus is or is not a person and thence – with some degree of innocence, and some degree of self-deceit – can treat it as a baby when it is wanted and as a piece of tissue when it is not. But this is irrational. The matter is one of life and death and one is put upon decision. Determination of the personhood or otherwise of the foetus is at once an intellectual impossibility and a moral imperative, it seems.

But the argument cannot be allowed to end in impasse. Since determination of personhood is impossible within our limited understanding of “personhood” (is a congenital idiot a “person”?) , and since we must regard abortion as yet raising moral questions of the first importance, argument centred on personhood may be misplaced. Ronald Dworkin (“Life’s Dominion”) seems to have a better approach: it is the failure to respect life that makes casual (or near-casual) abortion a moral crime. Life, all life, is deserving of reverence. The sheer wonderfulness of life, the vastness of the intelligence that raises life – to fail in reverence here is to fail oneself and is to indulge in a more or less deliberate blindness to realities and values. The more developed the form of life, the more creative endeavour it manifests, the greater should our reverence be, and human life or potential human life is the highest form of life we ordinarily encounter. This conclusion may seem to be unsatisfactory for vagueness and would not exclude abortion in at least some circumstances, while it would hugely complicate the choice, but the time may have come to abandon the simple certainty of the simple rule. Carnivores kill. Animals eat plants. They do so because they must. Killing is not wrong where it is done for grave-enough reason (though, as we shall see shortly, the problems begin there rather than end there).

And there is much that could be argued from other points of view, but this is not the place to serve up a re-hash of the abortion debate: the point of the above digression is to illustrate how moral choice strains the limits of human ability, of human thinking, how (in the present case) a formally indefensible distinction between direct and indirect intention manifests an attempt to grope a way towards moral truth. To return to the main argument: it looks as though the only meaningful distinction that can be drawn between direct and indirect intention, in relation to one’s own act, is the distinction between primary purpose or end, and (secondary) means or incident. This leaves it unclear why a sharp moral distinction should be drawn between end and means once the choice has been made in favour of end or purpose at the cost of means or incident. The moral question that arises is, rather: what is it that justifies the choice of this end at the cost of this means? Indeed, ends and means are not always sharply distinguishable from one another. Aquinas pointed out, long ago, the error that may be raised by doing so: means are themselves proximate ends and ends are proximate means to further ends.

The third condition is that the evil effect must causally be as immediate as the good effect. A variation of this is: the good and evil effects must be independent of one another so far as concerns any chain of causation (which of these ways of putting it is to be regarded as authoritative seems to be uncertain). Either way, the meaning is not clear. The condition assumes the validity of the principle of causation. Granted that principle, the notion of causal independence (the variation mentioned) is at least puzzling when one is considering a unified act, or a series of acts connected with one another in accordance with a purpose and performed towards that purpose. To return for a moment to the case of ectopic pregnancy: Roman Catholic teaching forbids the (simple) removal of an ectopic foetus on the ground that the evil effect (killing the foetus) is the sole immediate effect of the act. Removal of Fallopian tube is, however, permitted and where this is done that evil effect is regarded as having been produced only indirectly. But – so far as concerns the third condition – what decides what effect is “immediate”, and why? Either act has two effects: it kills the (doomed) foetus and saves the mother’s life. Which of these effects is “immediate” looks to be a matter that can be settled only by reference to intentionality and not by reference to any ‘bare’ act (and in any event, as we saw earlier, a human act cannot be ‘bare’ in the sense of being separable from intentionality). Again, a craniotomy in medically-indicated cases has two immediate effects: it kills the (doomed) foetus and removes a threat to the mother’s life. In the “just war” case, on the other hand, Roman Catholic teaching would permit, say, the bombing of an armaments plant. Yet any good effect here is anything but immediate. It is separated from the immediate act by an even longer chain of causation.

In few human acts are the good and bad effects morally (or even meaningfully) to be distinguished by reference to what is or is not causally immediate. Bruno Schuller has pointed out that, in vaccination, the evil effect (an artificially-induced infection) is immediate rather than is the good effect (immunisation). Yet vaccination is not forbidden by those who promulgate the principle of the double effect. No more forbidden are certain medicaments whose good effects (relief from distress, etc.) precede certain bad effects (such as a degree of dependence). Whether good effects and bad effects precede, accompany or succeed one another looks rather to be a matter of non-moral fact than a consideration to determine the moral status of an act. A unified act or series of acts (popping a pill, jabbing a needle into someone’s arm, removing the cancerous uterus of a pregnant woman) may have good effects and bad effects and it is unconvincing to assign moral status to such an act or series of acts by reference to temporal priority of effect.

The fourth condition goes to the heart of the matter: the evil that is foreseen and allowed may be allowed only for a proportionately grave reason.

But how could one settle this, save according to some prior scale of values according to which the good being sought through a particular act or procedure or undertaking was judged to be more important than the incidental harm? What settles that scale of values? How grave does a reason have to be, to be proportionate? In any event is the condition not equivalent to the rule: “choose the lesser evil”, and does not that (leaving aside the other conditions that are meaningless anyhow, apart from their attempted alignment with the Pauline principle) amount to a consequentialist ethic, and is consequentialism the thing to aim at? And indeed how does consequentialism differ from any deontic ethic? That is, on what deontological, intuitive or other basis does the consequentialist settle his scale of values in accordance with which a consequence of one kind is to be thought better than, or a lesser evil than, another? The idea of “proportionately grave reason” seems, so far from settling anything, to leave one still up to the neck in the fundamental problems of ethics. So far as concerns the Pauline principle and the principle of the double effect it is enough to note here that the idea simply has not been given meaning to an adequately workable degree, at all events not in any form that has received general recognition. The conclusion is more than somewhat momentous: the principles are empty. That is, apart from specific kinds of cases where authority has determined their application for those who accept that authority, the principles are of no help towards deciding what it is you ought to do when a moral problem arises.

If discussion is to be kept within reasonable compass, further remarks will have to be fairly brief. Recent discussion of the meaning of “proportionately grave reason” has led to nothing useful. A leading exponent, R. McCormick, begins by explaining it as meaning that the value that is being served should not itself be either attacked or undermined by what is done in its purported service, that is, that the principle that is being affirmed in the end sought must not be negated by the means adopted, or that means and end must not contradict one another. This sounds promising; however, it seems to imply that the life of the unborn child must never be sacrificed to save the life of the mother, and that the soldier should not throw himself on the grenade to save the lives of his comrades, and one worries that a purported application of the principle of non-contradiction should yield such conclusions. The point may be left unresolved because it is quickly overtaken.

McCormick cannot restrict “proportionate reason” to cases where end and means involve one and the same value. He recognises also that different human values are simply incommensurable – there is not, even in principle, a way of comparing different values so as to settle which is to be preferred when choice must be made.

“Give me liberty, or give me death”. Why? Should life count for more than liberty, or the other way round? Locke’s answer is fairly reasonable: the one who has you in his power is in a position to take everything that makes life even minimally worthwhile. But maybe he won’t take everything. Anyhow great numbers of people live without liberty, and some even manage to thrive, while believers in “the incomparable value of human life” are presumably not interested in quality-of-life questions. In the existential choice some fight and die rather than yield political liberty. Some don’t. Some practice contraception so as to be able to provide a better upbringing for fewer children. Some don’t. The practical choice gets made, because either it must, or mere confusion supervenes. But theory gives no help, nor does the principle of the double effect, and McCormick virtually admits as much. In the end his “proportionate reason” invokes just about everything that the careful and non-doctrinaire moral agent canvasses anyhow in an attempt to decide what to do: consequences (mediate, immediate, far-flung), the universalisability test (“what if everyone were to do this?”), the cultural climate of the time and place and the prejudices, reactions, aspirations that will be countered or favoured, the wisdom and experience of the past so far as mediated by rules, maxims, worthwhile opinion, the ordo bonorum …..

Casuistry (proportionate reason) is in important aspects a sophisticated form of consequentialism and takes account of matters that consequentialism seems to know little about.

One of the useful things that McCormick’s discussion brings out is that the idea that some acts are “intrinsically evil” is a mere mistake. Well, it is not a mere mistake: it is an error that has burdened human beings with an unnecessary load of guilt. Killing, lying, sexual activity may be right or wrong, depending on the circumstances. Yet (McCormick does not say this) the Pauline principle depends on this error, on the idea that the doing of something can be evil in itself and apart from motive, circumstances or other consideration. It is this error that raises the problems involved in the idea of doing “evil” that good may come of it, problems that the unsustainable principle of the double effect seeks to address.

A particular application of the idea that some acts are “intrinsically evil”, and one that gives rise to similar problems, is the proposition that there are certain “inviolable human rights”. It really does look as though some kind of psychological quirk may be the root of such manifestly erroneous notions. For if a person has an “inviolable right” to, say, life, then you may not defend yourself when he comes for you with an axe. Possibly such errors go back to the necessarily over-simplified moral norms with which most civilised persons are indoctrinated in childhood, possibly not. Possibly the idea of inviolable or absolute human rights is related to the Kantian prohibition of using a person only as a means. Whatever the provenance, one is put upon critical examination of received ideas when childhood has passed. All harm that human persons endure on earth, or inflict on one another, is by definition pre-moral or non-moral. In general, immoral action is action that inflicts harm (or fails to avert it) without good reason and the entire question turns on what is “good reason” in each particular case. If the discussion above of the distinction between direct and indirect voluntareity is correct in concluding that the distinction simply distinguishes what has been chosen as end from what has been chosen as means, then to use a person only as a-means is to inflict harm on him for the sake of what is taken, all things considered, to be a greater good – not his own good, but the good of another or of others. It is not self-evident that doing this is always wrong (McCormick makes this very point). For example, in so far as the sanctions of the criminal law are taken to be deterrent, a criminal is punished so as to deter others, whether or not the punishment does him personally any good, and even if he cannot himself be deterred (and in this case at least the harm is not inflicted on him until, by some criminal act, he has forfeited his right to personal inviolability). The legitimacy or otherwise of lying to the intending murderer may be judged rather by reference to a right (not to be lied to) that the intending murderer has forfeited them by reference to an attempted comparison of incommensurable evils (lying, or failing to protect the life of another). The legitimacy of lying to the intending murderer in order to protect even a quite minor interest of one’s own could at least be argued on the same ground, viz: the intending murderer is an “outlaw”. He has voluntarily stepped outside the world of law and order and is entitled to none of its protections / rights – although this might be to generalise unduly.

All this leaves little enough to be said about the Pauline principle proper. The harm that is not to be done, that good may ensue, is “evil” if it is not proportioned to that good. Otherwise it is not “evil”. And moral evil as such is not to be done anyhow: the context suggests (though uncertainly) that that was what Paul was actually driving at.

What is wrong with the Pauline principle – when an attempt is made to interpret it literally – comes out in Alan Donagan’s discussion of it (“The Theory of Morality”). Donagan claims (pages 154 – 5) that the principle means that a positive duty (help others in need, develop oneself, etc.) may not be satisfied by way of breaching a negative duty/prohibitory precept. Wouldn’t it be nice if things were so simple! Take an example: Donagan assumes (p.167) that sterilisation (of a human person) is absolutely prohibited. But if means of contraception are for any reason not available, how then can family size be limited? Donagan ducks: he remarks (p.168) that “the disastrous overpopulation that afflicts … societies like India and Bangladesh seems to have originated in avoidable failures of understanding and will” and follows with one or two other general remarks that likewise simply do not bear on the question: what, exactly, is a poor Bangladeshi family to do? Another example is the often-discussed case of the stuck potholer (who must be dynamited out, and killed, if his trapped companions are not to drown). Donagan’s solution (pages 177-180) does not rest on traditional morality at all, but on implied contract. That is, the potholers, voluntarily engaged in a hazardous enterprise, may be taken to have agreed as to what should be done in such a crisis -though Donagan avoids admitting that contract here displaces the moral norms and merely claims (without showing justification) that such contracts (involving one’s consent to being deliberately killed in certain circumstances) are consistent with common morality. Doubtless this is news to some traditional moralists.

Or again: Donagan takes it that self-mutilation is, like suicide or sterilisation, barred by the prohibitory precepts. Yet he admits the legitimacy of surgical operations “for the sake if one’s well-being as a whole” and appears simply not to notice that in doing so he has jettisoned the absoluteness of those prohibitions. If surgery, why not also vasectomy or tubal ligation (for the sake of the well-being of the family as a whole) in a case where his “principle of culture” counter-indicates another child? On Donagan’s own principle that the perfect (or negative) duties bind absolutely while the imperfect duties bind conditionally, surgery (i.e. self-injury, a breach of a perfect duty) cannot be justified by reference to an imperfect duty concerning the promotion of one’s well-being. Again, he defends (p.80) the (moderate) use of intoxicants, and of drugs such as tobacco, “inasmuch as the relief and enjoyment afforded by a drug compensate for any ill-effects it may have” – a clear breach of the Pauline principle.

As regards Kant’s question about lying to the intending murderer so as to protect his intended victim, Donagan at one point (p.89) approves of the lie and at another (pages 146-148) appears to disapprove of it. Yet again: “One may not blind oneself to save another from blindness” (p.79) although he has just (p.78) agreed that Captain Oates was not wrong to kill himself in his attempt to save the other members of Scott’s party. The examples could go on. However, Donagan is by no means alone in these confusions. His book is taken simply because it is a fairly recent and fairly thorough exposition of traditional morality.

The absoluteness of the prohibitions traditionally taken to be absolute, and the principles founded on that absoluteness -the Pauline principle and the principle of the double effect – rest on the conviction that what absolutely must not be done is an immoral act (injustice, malice etc.). How could anyone not agree, who takes the human person to be more than his earthly embodiment, and how could anyone have ground for caring one way or another who doesn’t? The mistake arises in identifying certain physical acts or omissions as unjust or otherwise immoral in themselves. The Pauline principle can be invoked where alternatives (even thoroughly unpleasant alternatives) have not been canvassed to exhaustion, but this can be of limited assistance because the alternatives must then be evaluated, according to a judgement of some kind, unless simplistic “wrong in itself” is to rule. A breach of obedience to caritas may reasonably be held to be wrong in itself, and yet cannot categorically be identified with particular kinds of acts.

The continued defence of those principles is a doubtful service to morality. It is at least possible that outsiders see in that defence something rather less pleasant than honest mistake. It is possible that they suspect a determination to invoke any argument, use any means, to uphold a position. If they do they may simply walk away, as so many have walked away from the organised Churches, believing that this disputant is interested, not in seeking out the truth, but simply-in putting across his own view. Statements that begin: “It is absolutely forbidden to…” begin to look rather like attempted bullying (though perhaps they represent only attempts at self-reassurance) once one finds how carefully one must search in order to bring clearly to consciousness what really is absolutely forbidden, how deadly serious a matter it is, and how difficult if not impossible it may be to decide with any certainty just whether a particular act in particular circumstances is or is not forbidden. Donagan says (p.88) “… it is absolutely impermissible to commit murder. However, the precise scope of the class of the innocent is still in dispute”. So what is it that is absolutely impermissible? He is at least admitting to the difficulty, or is almost doing so, even though he elsewhere denies that moral dilemma simpliciter can arise under the rules of traditional morality.

Every concrete decision that falls to be made in the course of actual living, every situation in which some kind of loss or harm will result from any available choice, comes down to the question of having sufficiently good reason for doing or not doing, for doing this or doing that. That “good reason” may engage and draw upon everything you have and are, your whole being. By all means take out your consequentialist abacus, but calculation of consequences is only the beginning of it. What preferred consequences, and why? What rights are recognised, and are yet to be overridden, and why? Will others be strengthened or weakened by what they think they see being done? What alternatives are available at what cost to whom? How much good (to someone), how much harm (to someone), what are the claims of each of these in relation to one another, to the moral agent, to the world at large to justify this benefit to one, this harm to another? Are the claims in this case grounded primarily in contract? In the good of order? In caritas? In the judgement of experience, the traditional rules?

The gap between a teleological ethic and the obligations of justice / fidelity / fairness / gratitude / reparation may, sometimes, be closed by attending carefully to the consequences to be sought, and why those consequences. Good consequences are sought in the service of caritas (no other rationale seems plausible). But caritas includes rationality. Rational caritas insists that the requirements of justice, fidelity etc. are so important as usually to count for more than material consequences, because obedience to those requirements is what constitutes good inter-relationship. Harmony, beneficent inter-relationship, is the substance of caritas. Caritas cannot be served if the ‘good of order’ is not served. Justice, the good of order, forbids the sheriff to hang the innocent man. Even here one should perhaps hesitate: it could be intellectual arrogance to claim that no case could ever arise to justify any other course. We don’t know what cases may arise in the future. For the moment we do our best and stick to the rule: you must not hang the innocent man, while yet recognising that in principle the moral agent cannot shuffle off his responsibility to rule-obedience.