The Lay-out of Arguments

By J. R. Lucas, Oxford

from W.Krawietz, N.MacCormick, G.H.von Wright, eds.,
Prescriptive Formality and Normative Rationality in Modern Legal Systems,
Festschrift for Robert S. Summers
Berlin, 1994, 285-295


Arguments have been much misunderstood. Not only has it been assumed that they must be deductive, but it has been assumed also that, although often expressed in loose and elliptical form, they must be capable, if they are valid at all, of being expressed with absolute precision, as a set of necessary and sufficient conditions for the action in question to be appropriate. Mathematical arguments are capable of being stated precisely, and it has long been a reproach to workers in other disciplines that they do not manage to achieve equal precision in their work. The assumption is made that absolute precision is in principle available, and it is only a lack of rigour by practitioners in the non-mathematical disciplines that prevents them casting their arguments in satisfactory form, and the philosopher should, therefore, reconstruct arguments to conform with these requirements, evaluating those that do, and rejecting any that do not.

This assumption is false. Most arguments, not only about morals, but in the humanities generally, are of a different, dialectical, form. A case is made out prima facie and objections to it are made. These objections may be rebutted, and counter-arguments countered. Sometimes the argument may be conclusive, so that there is no more to be said, and often some consideration is decisive so that we have little doubt in making up our minds. But often we do so leaving open the possibility that further considerations could emerge which would justify a change of mind, and often we reach a conclusion only tentatively, saying that the facts support it ceteris paribus, other things being equal.1

It is clear that there are very many variations on this pattern of argument. We may object at the outset to the prima facie case so that it never gets off the ground, or we may shoot it down in full flight, or we may allow that it strikes home but still claim to escape its full force. These different counters may be met by different rebuttals, and the argument may be ramified in different ways. The possibilities are extremely complex. Searle criticizes Ross for not

1. W. D. Ross, The Right and the Good, Oxford 1930, pp. 19ff.; H. L. A. Hart, "The Ascription of Responsibility and Rights," P.A.S. 1948, pp-171-94; reprinted in A. G. N. Flew, Logic and Language, vol. 1, Oxford 1951,pp. 145 - 65; J.R.Lucas, "The Philosophy of the Reasonable Man," Philosophical Quarterly, 1963, pp. 97 - 106; J.R.Lucas "Not 'Therefore' but 'But"' Philosophical Quarterly, 1966, pp. 289 - 307; S.E.Toulmin, The Uses of Argument, Cambridge 1958, pp. 57ff.


286 J. R. Lueas

distinguishing them sufficiently,2 and, indeed, there is a difference between the case where I argue "You ought to go to Sheila's party because I heard you say to her 'I hereby promise to go to your party" which you counter by saying "You should not pick up fag-ends: I only uttered those words as a philosophical example", and the case where my argument is countered "But I have had a telephone call saying my mother is ill, and I must go at once". In the former case there never was an obligation. Although your having uttered the words is prima facie evidence of your being under an obligation, the circumstance of their being uttered as a philosophical example entirely alters the case. So, too, if you were play-acting or sleepwalking. Other things being equal, to say 'I hereby promise to go to your party' creates an obligation: but other things are not equal, and the prima facie case is defeated by the special context of utterance. In the latter case, however, there was an obligation, only it was overridden by another more fundamental, more pressing, more important or more urgent obligation. An overridden obligation is not the same as no obligation or an extinguished obligation. If you had never promised to go to Sheila's party, or if you had promised but had subsequently been released from your promise, there would be no call on you to explain your non-appearance or present your apologies or make amends. If, however, you had promised, but the obligation had been overridden by another, then you would need to make it up with Sheila. Although you had done the right thing in going to your mother, you had failed to do what you had promised to do, and you owed Sheila an explanation. Once given, it should, of course, be accepted; she should lose no time in acknowledging your greater obligation to your mother, and expressing sympathy for you at a worrying time. But the fact that you owed her an explanation shows that the obligation was not extinguished by being overridden - as it would have been if she had cancelled the party -, and often the obligation remains in attenuated form - the next time she asks you to come you ought to make a special effort to be able to, even at the cost of considerable inconvenience.


Thus far Searle is right to warn against the dangers of oversimplification. But this scarcely diminishes the importance of the concept of the prima facie that Ross introduced. At each stage an argument is adduced which is not conclusive but is adequate. In normal circumstances, it is enough that I heard you say 'I hereby promise to go to your party' for you to have put yourself under an obligation: but there can be circumstances which render such an inference incorrect. I do not need to ascertain in advance that no such circumstances obtain. I am entitled on the evidence in my possession to draw that inference. But it is defensible, and can be defeated should some exceptional circumstances obtain. If, however, on enquiry no such circumstances can be discovered, then

2. J. R. Searle, "Prima Facie Obligations", in Joseph Raz, ed. Practical Reasoning, Oxford 1978, pp. 81-90.

The Lay-out of Arguments 287

I may reckon that, having considered all the circumstances of the case, you really did put yourself under an obligation. Similarly, if once it is admitted that yourself are under an obligation, the inference can, other things being equal, be drawn that the obligation still holds. Such a presumption needs no further evidence to support it. But it can be defeated - if Sheila subsequently released you from the obligation, or said or showed that she did not reckon it to be a firm promise, or did not expect you to come or did not really want you to come; or if war breaks out, or she gets ill, or the university is closed down by the Minister of Education. In the absence of such special circumstances, the inference is valid, and you are under an obligation. From this it normally follows that you ought, indeed, to go to the party when the time comes, and that only by doing that will you have acquitted yourself well as man of your word. But other duties can arise, and in view of them you may conclude that in the event you ought to do something else. It seems quite reasonable to call the obligation to go to the party a prima facie obligation to distinguish it from what, after due consideration, you conclude you actually ought to do. Such a prima facie obligation may be overridden by a more important or more pressing one, and in that case you should carry out the latter with a good conscience and without repine, although, as we have seen, with due explanation and apology. There is a difference, which Searle is right to insist on, between this latter case and the others, because prima facie obligations on our part give rise to legitimate expectations on the part of others, and if we have to disappoint them, even for good reason, we still need to make amends. But in spite of this difference, there is an important similarity between the pattern of inference at the last and at earlier stages of the argument, and Ross was right to see that moral arguments, as well as many others, needed to be characterized in a quite different way from what had hitherto been assumed.

The dialectical nature of practical reason has important consequences. Even when there is no moral conflict, there are often reasons for an action and reasons against. Some of these reasons may meet others and defeat them, but often we shall be left still with reasons on both sides, and will have to decide between them. We deliberate and try and assess how weighty the reasons are on either side, and finally strike a balance. As the metaphor suggests, the decision is not an arbitrary one; but nor is it governed by some impersonal decision-procedure. We can criticize someone's decision, but we cannot replace it by the result of a standard application of an impersonal decision-procedure. It is a decision which has to be taken by somebody or other, using his judgement as best as he can. We may disagree with his judgement, and reckon that had the decision been ours, we should have decided differently. But then it is our judgement against his. We may be right, but we are not, save in exceptional circumstances, indisputably right. From this it follows that we cannot brush aside the reasoning of a decision-maker, even when we dissent from his decision, in the way we can ignore the calculations of a faulty arithmetician, the

288 J. R. Lucas

fallacious proofs of a would-be geometer, or the chemical analyses of an incompetent experimentalist. With moral argument - and many other forms of argument in the humanities - we have to enter into the reasoning of the man who strikes the balance, and not merely replace it by what we take to be the correct reasons. We have to consider not merely the reasons, but his reasons, and take not merely a third-personal, or rather omni-personal, standpoint but also a first-personal standpoint as well.

The interplay between first- and third-personal reasons is of great importance in moral philosophy. While we are deliberating we consider third-personal reasons for and against a particular course of action. When we have decided, we support our decision by reference to the reasons for or the reasons against, whichever are the reasons inclining towards the way we have decided. When asked to justify the decision we give our reasons, namely those reasons which in our judgement weighed most heavily with us. In deliberating we weigh the reasons on either side against each other, and in striking a balance we adopt the weightier as our reasons, which we cite thereafter in justifying our decision. We do not cite the reasons on the other side, but only our objections to them, if they are pressed on us. Although while we were deliberating we weighed them, they did not weigh with us in the event, and drop out of sight in our subsequent report of our reasoning. In discussing what is to be done, you and I may well agree on what are the relevant considerations on either side, but strike the balance differently, so that I decide one way giving as my reasons some of the considerations inclining one way and you decide the other way giving as your reasons some of the quite different considerations inclining the other way. In debate we largely agree on what the relevant considerations are, but in conclusion we cite entirely disparate selections from those considerations in support of diametrically opposed decisions. It is easy, then, in retrospect to concentrate on the difference of judgement while overlooking the similarity of consideration. And this makes our first-personal reasons seem more subjective and less inter-personal than they really are. Real differences of opinion mask considerable agreement in argument.

Suppose we are arguing about capital punishment. We agree that life is sacred, and that it is wrong to kill. One takes this as an absolute, and concludes that judicial execution is therefore wrong. The other contends that those who have killed have forfeited their right to life, and may be condemned to death to prevent their ever killing again and to deter any others who are tempted to kill. The debate widens. The advocate of the death penalty points out that even the abolitionist does not regard the prohibition of killing as being absolutely absolute, and will allow that a man may kill another in self-defence. The abolitionist points out that imprisonment is as effective a preventative as execution, and claims that it is also equally effective as a deterrent; and he emphasizes the possibility of error in the judicial process. In the debate each

The Lay-out of Arguments 289

takes the points made by the other. Both agree on the sanctity of life - the murderer's, his victim's, and other potential victims'. Both agree that wrongdoing merits punishment, and that crimes should be prevented and discouraged. Both agree that miscarriages of justice are very terrible, and not to be tolerated or made light of. But in giving their reasons for holding opposing opinions each will cite only those considerations which he finally takes as decisive, and will ignore the large measure of common ground. The advocate of the death penalty will instance the rights of victims and potential victims as paramount, the abolitionist the importance of the state's not compromising by its actions its stance on the absolute sanctity of human life. Their reasons - the reasons each adopts as his own in coming down on one side or the other - differ: but the reasons they recognise as relevant in debate are largely the same.

The dialectical structure of practical reason has the further consequence of making it informal. Few formal rules are applicable where everything dep6nds on the response the other party makes. If I adduce an argument you may counter by denying the facts I allege, or by glossing them further - "although I did indeed utter the words 'I promise to pay you five dollars' I did so only as a philosophical example and not in order to make a genuine promise" -, or by denying the principle of morality I invoke, or by claiming that it does not apply in the particular case in question, or by acknowledging the force of my contention but claiming countervailing considerations which override it. My next move will depend on which move you make, and your counter on mine. There will thus be a rapidly widening variety of moves and countermoves, and no simple pattern that comprises them all. At each stage it will be necessary to consider the substantial point that has been made, and to address one's mind to that. No formula will fit all the twists and turns of dialectical argument, and we have simply to argue as best we can on the substantial merits of the case.

There are many different counters that can be made to an argument, but not all of them will be made. Although it is logically appropriate to respond to an argument based on the fact that I uttered the words 'I promise to pay you five dollars' by explaining that although I did, indeed, utter the words, I did so only as a philosophical example, it would often be untrue to make such a response, and often I concede that I not only uttered the words but did so with the intention of their being taken in the normal way. I concede, as in all honesty I must, many points, and concentrate on the crucial ones where my counter-argument can be substantiated, and if substantiated will carry the day. There is thus an important gap in the many points I do not contest between the degree of proof actually required and the degree that might on some occasion conceivably be required. Often it is enough that you heard me say 'I promise to pay you five dollars', but on some occasions we can conceive of we should have to consider the possibility of the words being uttered not in order to make a promise but to serve as a philosophical example, and then we should need to

290 J. R. Lucas

consider the whole context of the utterance. This possibility, however, does not lead to our always requiring evidence of context before we are justified in taking the words 'I promise to pay you five dollars' in the standard way. Once we are reliably informed that the words were uttered we are entitled to assume that they were intended to be taken in the standard way unless reasons are adduced for believing the contrary. Moral arguments do not proceed more geometrico. A geometrical argument is no argument at all unless it is absolutely watertight: a moral argument, together with most arguments in the humanities, is full of possible holes, but nevertheless holds water unless someone succeeds in picking an actual hole. A typical "proof" in morals, history or literary criticism is one which gives adequate reason for the conclusion in the absence of a convincing counter-argument, whereas a proof in geometry or mathematics generally is one which gives conclusive reason, a logically sufficient condition, for the theorem, with there being no logical possibility of any counter-argument whatsoever. Much therefore depends in moral argument on a possible hole not being an actual hole, on the other party's conceding the point rather than contesting it, on his letting it pass in silence rather than interposing a further 'but'. The argument from silence is of great importance in morals, as it is, in a different way, in history.

It might seem that the most effective strategy in argument was to contest each and every point made by the other side, so as to require them all to be proved up to the hilt: in this way one would maximise the chance of finding some weak points which would unhinge the whole argument. Many children and some inexperienced advocates adopt such a strategy, but it is less effective than they suppose. For often it is a matter of judgement whether an objection is cogent, and if someone shows bad judgement in putting forward objections which are not sustainable, it weakens the authority of his judgement over all, and in particular about those objections which might be sustainable. If I make out that when I uttered the words 'I promise to pay you five dollars' I was using them only as a philosophical example, I put you to the trouble of proving that I used them in the standard way: but if you succeed in doing this, I am seriously discredited, and my other arguments are correspondingly weakened. I am much more likely to be believed if I say simply that I had in fact discharged the debt, or that you had subsequently released me from my obligations, and do not also make out something that turns out to be demonstrably false. If I am wrong about that, as I evidently am, I may well be wrong on the other point too. So it is that in morals many reasons are often weaker than one: whereas nobody thinks Pythagoras' theorem less true because it can be proved in forty-seven different ways.

The dialectical structure of moral arguments reflects the fact they are practical arguments - arguments about what ought to be done - addressed to human beings, who are rational agents acting under conditions of imperfect informa-

The Lay-out of Arguments 291

tion. Because we are human, we are fallible; because we are having to decide what to do under conditions of imperfect information, we have to live with our fallibility, and decide as best we can none the less. We cannot not decide. In the practical world of affairs, tertium non datur: either one does something or one does not; if one does not accept the invitation by the time the party comes, one has declined it. It is a rare luxury, mostly confined to academic life, to be able to suspend judgement. Usually we have to decide on the evidence available at the time, even though the evidence is inadequate and our judgement faulty. We have to make our minds up what to do, and shall be lost if we hesitate too long. I cannot dither, but must decide either to stand and fight, or to turn and flee while there is yet time. My decision may be wrong, and so there is always room for a further 'but' between the arguments and the final conclusion, allowing that the arguments were not bad but pointing out that in the event, in the light of some further factor or some further consideration, the conclusion was in fact the wrong one. Practical arguments are full of possible holes. But although it is inherent in taking decisions that we can be wrong, it does not follow that we cannot be right. We are rational agents, acting as best we can in the circumstances in which we find ourselves, and although we may not succeed in selecting the right response to the situation, some responses are more appropriate than others, and we may, and indeed often do, succeed in responding appropriately. So, although we can be wrong, there is a presumption of rightness. If a man judges something to be the appropriate response to a certain situation, the fact of his so judging is an indication in its favour. He may be wrong, but does not have to prove he is right. If we doubt the wisdom of his action, we may ask him his reasons, and may then fault them or point out reasons bearing the other way: but it is for us to make the counter-argument; he does not have to give his reasons and then go on to prove that they are good ones. Although it is theoretically possible, and may occasionally happen, that someone adduces a completely bizarre reason for an action or a moral obligation - "You ought to pay me five dollars", "Why?" "The number of the Beast in the Apocalypse is equal to the height of the great Pyramid measured in cubits" - to which the only response is a request for further explanation why such an alleged fact should constitute a justification for the obligation's obtaining, these cases are extremely rare. In nearly all cases the relevance of the reason is not in question. The once popular riposte "So what?" is nearly always inappropriate. Nearly always it is evident that the reason adduced is indeed a reason, as we should expect of a rational agent. It is very seldom that we do not recognise it as being a reason at all: rather, we question whether it is a good enough reason, or whether it is, although a good reason, overridden by other weightier ones, or whether, although in general a good reason, it is not inapplicable to this particular case.

Arguments between fallible rational agents are characterized by a good measure of disagreement but a greater one of mutual understanding. Because

292 J. R. Lucas

we are fallible, each is liable to error, and even if you are in fact right, I may be wrong, and so you will have to disagree with me. So we often disagree. But I can nearly always understand your reasons and enter into them, even though they are not for me the ones I adopted as the decisive ones. Hence on any particular issue we well may disagree, although we do not have to, but in articulating our reasons and adducing our arguments we operate against a background of mutual understanding and potential agreement. This is why argument is worthwhile. At the very least it engenders understanding. And by pinpointing the issues on which we disagree it reduces the area of disagreement and may lead to a complete resolution.

The dialectical structure of argument is not peculiar to moral argument. It is characteristic of all practical argument, and of all the humane disciplines, such as history and literary criticism, which are about human actions and reactions. In each case the fundamental concern is with the response of a human being to a set of circumstances. We make some deep metaphysical assumptions of shared humanity, believing that other people are in some deep respects likeminded with ourselves, but do not identify with them completely. Rather, we submit them together with ourselves to the impersonal ideal of rationality. We do not seek simply to understand, but also to judge or evaluate. Reasons are not simply first-personal reasons why the agent did or does or will do a particular deed, but also why he should have done it or should do it. We can identify with the agent to the extent of understanding why he did what he did or will do what he plans to do, and yet dissociate ourselves from him in reckoning his decision wrong, and in believing that in his shoes we should do differently. But by the same token we have to sit loose to our own opinions, and recognise that we too could be wrong, and submit our own views to criticism and possible correction. It is always difficult, notably in history, to combine the sympathetic with the critical role, and the critical with the self-critical. Historians have learned to be wary of the judgemental role in which the Kings of Judah are judged simply on the score of whether they did good or evil in the sight of the Lord. Tiberius, Caligula and Nero may have been bad men, but the historian hesitates now to anticipate the Day of Judgement, and is concerned to explain, rather than to evaluate, their actions. He seeks to understand why each acted as he did, to get inside him, and understand the action from inside. But that is impossible except on some assumption of common humanity. I cannot get inside Caligula's action in making his soldiers pick up sea shells unless I can see myself so circumstanced that I might be inclined to act similarly. It remains utterly opaque, the unintelligible actions of a madman, unless some circumstances are sketched in and some rationale suggested,3 which makes sense of the action. Then, and only then, I can see why Caligula acted as he did, because I can see that perhaps I should have acted similarly if similarly

3. As by J.P.V.Balsdon, The Emperor Gaius (Caligula), Oxford 1934, pp. 88 - 95.

The Lay-out of Arguments 293

situated. Should: I see as something which perhaps I should have done, something which perhaps was the reasonable thing to do; it is only by projecting myself into his situation, and considering my rational responses, that I can see things from his point of view; his first-personal reasons are accessible to me only because they are reasons, which can guide not just his actions alone but those of any rational agent, myself included. But perhaps: I do not have to adopt his reasons as my own wholeheartedly; it is enough that I can envisage that perhaps I should have done as he did, even though in his actual situation I should not have done so in view of some of the circumstances or some further considerations which he evidently did not take into account; perhaps, if those circumstances did not obtain or those considerations did not apply, I should and would have done as he did, and so I can appreciate what was behind his action even though in his actual circumstances and in view of all the considerations that applied, I should not and would not have done as he did.

Many modern historians, in their anxiety not to be judgemental, have failed to see the extent to which we must evaluate in order to explain. They suppose that all evaluation must be wholehearted, and must lead us simply to praise well-doers and blame evil-doers, and in order to avoid that, they make out that history must be value-free. But it is, as we now can see, a false alternative. The Whig ideal of history is not the only one that allows evaluations. Granted the dialectical structure of practical argument, I can see why an agent acted as he did, and enter into his reasoning and understand the action from the inside, without wholeheartedly endorsing it. I can see that he had his reasons for acting, and recognise that that they were indeed reasons why he, or anyone, should do as he did, and yet interpose a mental 'but', citing other reasons bearing the other way, which would lead me to refrain from acting as he did. I can evaluate without endorsing, and can therefore explain and understand without needing either to condone or condemn. But, of course, I may be led to justify or condemn. In understanding the actions of Caligula, I may see why he acted in a certain fashion, and reckon that it was indeed a reasonable thing to do, and that he was, in this instance at least, unfairly traduced by Tacitus. In understanding his action, I may endorse it; although I may dissociate myself, interposing a further 'but', I may not, and may be led not only to appreciate his reasons but to appropriate them as what would have been my own; equally, I may dissociate myself, and reckon that although Caligula may have had some reasons, which in some circumstances might have led me to act similarly, in the actual circumstances of his case there were overwhelming reasons the other way, and his act was an act of folly or of great wickedness. Although to sympathize is not the same as to criticize, it carries with it the possibility of criticism. I cannot see things from someone else's point of view without entering into his position; but it is I who am entering in, and in evaluating his action from his point of view, I cannot help being in a position to evaluate it from mine. If I were to forswear all possibility of judging and criticizing, I should

294 J. R. Lueas

have to forgo all evaluation and explanation. The one carries with it the possibility of the other. But although I have to be able to justify and condemn, I do not have to be doing so all the time, or to make it the main task of the historian. The Whig interpretation of history sees it as a progress towards the Whig's present situation. But history need not be interpreted as a progress, and the present need not be seen as the end to which all things should move. The historian can be slow, rather than quick, to criticize or judge historical characters from his own standpoint. Although necessarily occupying his own standpoint, he may remember how different past ages were from his own both in their circumstances and in the moral principles inculcated into those living then and in the moral standards expected of them. If I were running the Trojan war, I should not act as Agamemnon did, or so I hope: but if I had been brought up as Agamemnon was, and had not read the dialogues of Plato or the prophets of Israel, and knew nothing of the civilities of Cicero or the teaching and self-sacrifice of Christ, can I be sure that I would not have been as aggressive and self-assertive as he? A certain degree of charity is appropriate in view of my inadequate appreciation of the whole situation in which he found himself. I may be able to put myself into his shoes, but can never put myself entirely into his skin, and so should be cautious in criticizing. Quite apart from this, the historian may hesitate to judge knowing that he, too, is subject to judgement, and that his own opinions and the moral principles of the present age are not necessarily correct. There is a certain self-righteousness about the Whig interpreter of history that we do well to eschew. Although in explaining we are in a position to criticize and on occasion must, we do not want to obtrude our judgements unnecessarily, any more than we do in the day-to-day business of our present age. I must on occasion pass judgement: it would be inhuman to recount all the actions of Caligula or Nero or Hitler or Stalin in an entirely neutral, value-free way. But our purpose in writing history is often to explain rather than to condemn, and so we are not eager to mark what has been done amiss.

Literature holds a mirror to life in a different way from history. It dispenses with the condition of actuality with all its consequential clutter. The poet, the playwright and the novelist portray possible people in possible situations responding in characteristic ways where adventitious circumstances do not obscure the plot or obstruct the course of poetic justice. Even more than in history we understand by entering in, seeing the situation from the agent's point of view, and being moved and responding in the same way as he. Literary criticism is one remove from literature, and is concerned with other things - the magic of words - besides the portrayal of character and the development of action. Nevertheless, much argument in literary criticism has the same dialectical structure as historical argument. An interpretation is put forward, and a prima facie case made for it by reference to various passages. Objections may be made on the basis of other passages, and these may be countered.

The Lay-out of Arguments 295

Alternative interpretations may be suggested, which may or may not commend themselves in preference to the one originally proposed. What considerations are relevant will be largely agreed between the critics, but they may well differ in the assessment of where the weight of argument lies. In one respect there is an important difference between discussing a work of art and deliberating on a course of action. At the end of deliberation we have to decide, choosing one course of action and rejecting others: interpretations are not similarly exclusive, and although we may reject an interpretation on the ground that it seems inadequate, we do not have to on account of our seeing merits in another. Both interpretations may be illuminating, and if we find them so, we may accept them both.

Not all humane, not all practical arguments are moral arguments. We may confine our attention simply to understanding a man's action or a work of art, deliberately suspending judgement on it: or we may in our deliberations confine our questions to what is expedient, what is prudent, what is profitable, or what is socially or legally obligatory. In asking what I should do, I may exclude certain considerations - I may exclude consideration of the interests of others, or consideration of my own; I may be concerned solely with my duties as a citizen, or as a tutor, or as a don; I may look only to future consequences, or to past commitments. These exclusions and emphases colour the arguments, and the conclusions obtained. The tenor of some is moral - altruistic, utilitarian, or deontological: others we contrast with morality, and call them counsels of prudence, or reasons of state, or considerations of self-interest. But although the course of these arguments all differ, they all have the same general lay-out, and exhibit the same underlying dialectical structure.





Return to home page