Historical Notes 012, History and Human Responsibility (1990), by Professor Antony Flew

Share this



Historical Notes No. 12

ISSN 0267-7105
ISBN 1 85637 012 7

An occasional publication of the Libertarian Alliance, 25 Chapter Chambers, Esterbrooke Street, London SW1P 4NN

www.libertarian.co.uk email: admin@libertarian.co.uk

© 1990: Libertarian Alliance; Antony Flew.

Antony Flew is Professor Emeritus of Philosophy at the University of Reading. He has contributed countless articles to both philosophical and political journals, and his many books include An Introduction to Western Philosophy, Hume’s Philosophy of Belief, Crime or Disease, A Rational Animal?, God and Philosophy, Sociology, Equality and Education, Thinking About Thinking, Thinking About Social Thinking, The Politics of Procrustes, and Equality in Liberty and Justice.

This article was first published in the Rationalist Annual of 1962, and is reprinted with permission. The views expressed in this publication are those of its author, and not necessarily

those of the Libertarian Alliance, its Committee, Advisory Council or subscribers.

Director: Dr Chris R. Tame Editorial Director: Brian Micklethwait Webnmaster: Dr Sean Gabb


In the academic year 1960/61 Professor E. H. Carr gave the Trevelyan lectures in Cambridge. The essence of these was later distilled on the Third Programme as a series of six talks under the general title ‘What is History?’ The full original course was due to be published by Macmillan some time before this appeared. These reflections on his trade by one of our foremost historians can be seen as a major event in the rationalist year. For Carr’s outlook is this worldly through and through, and his approach is straightforwardly rational.

All this comes out very clearly in the talks. In the second, with an ostentatious cultural detachment, Carr comments upon the Robinson Crusoe myth: ‘Robinson is not an abstract individual, but an Englishman from York; he carries his Bible with him and prays to his tribal God.’
In the sixth, after criticizing one of Popper’s theses, Carr pays him a compliment as proper as it is characteristic: ‘On one point,  indeed, I should pay tribute ... He remains a stout defender of reason, and will have no truck with past or present excursions into irrationalism.’2 Much more important than any such asides is the fact that the whole of Carr’s final talk is concerned with ‘The Widening Horizon’, which is provided by the extension of the possibilities of reason in the modern period. He sees the work of Hegel and Marx, of Malthus and Freud, as a succession of landmarks in the always uneven progress of our attempts to understand men and societies; a progress which constantly opens up new possibilities for the planned and conscious change of both ourselves and our social institutions.

Carr insists that this widening horizon is one of the most important and most exciting features of the contemporary world. The point which he is making is, of course, independent of demurring questions about the exact scientific status of the particular theories of some of these intellectual heroes. It is made, perhaps, most forcibly where Carr dismisses the ‘popular charge against Freud, that he has extended the role of the irrational in human affairs’. This, he urges, is ‘is totally false, and rests on a crude confusion between recognition of the irrational element in human behaviour and a cult of the irrational.’ (It was, he might have added, Jung and not Freud who appeared willing to become High Priest to such a cult.) ‘What Freud did was to extend the range of our knowledge and understanding by opening up the unconscious roots of human behaviour to consciousness and to rational inquiry. This was an extension of the domain of reason, an increase in man’s power to understand and control himself, and therefore his environment; and it represents a revolutionary and progressive achievement.’3


But this vision, which we share, would surely be no more than a grotesque delusion if it really were the case that whatever happens in history happens inevitably. In his fourth talk, on ‘Causation in History’, Carr considers the question of how the possibility of historical explanation, which seems to presuppose some sort of determinism, is to leave any room for human responsibility, which requires that at least some of our actions could have been other than they were. This is a problem which in this particular form seems to have been raised and treated first by Hume. Appropriately enough, since this philosopher early formed and later fulfilled the ambition to become also an historian. Though Carr makes no mention of Hume, his attempt to resolve the dilemma proceeds on Humean lines; and is inadequate in at least the same ways.4

Carr directs his remarks in the first instance against Sir Isaiah Berlin’s well-known harangue on Historical Inevitability (OUP, 1954). This is a diatribe against historical determinism, which Berlin finds objectionable because, as Carr rather mischievously puts it, ‘by explaining human actions in causal terms, it implies a denial of human free will, and encourages historians to evade their supposed obligation to pronounce moral condemnation on the Charlemagnes, Napoleons, and Stalins of history.’ Carr defines determinism as  ‘the belief that everything that happens has a cause or causes, and could not have happened differently unless something in the cause or causes had also been different’. He then begins, like Hume, by urging that: ‘Determinism is a problem not of history, but a problem if it is a problem of all human behaviour ... Everyday life would be impossible unless we assumed that human behaviour has causes which are in principle ascertainable.’ (An aside compares ‘Sir Isaiah Berlin’s objection to our explaining why human beings acted as they did, on the ground that these actions are governed by the human will’, to the old obscurantist idea that it would be blasphemous ‘to inquire into the causes of natural phenomena ... obviously governed by the divine will.’5)

Carr proceeds to consider the historian. ‘Like the ordinary man, he believes that human actions have causes which are in principle ascertainable. History, like everyday life, would be impossible if this assumption were not made.’ As for the idea of historical inevitably, this is apparently all a mistake, the product of occasional careless talk. ‘Historians, like other people, sometimes fall into rhetorical language and speak of an occurrence as “inevitable” when they mean that the conjunction of factors leading one to expect it was overwhelmingly strong.’ But really the term inevitability is entirely redundant. ‘Nothing in history is inevitable except in the formal sense that, for it to have happened otherwise, the antecedent causes would have had to be different. As a historian, I am perfectly prepared to do without “inevitable”, “unavoidable”, “inescapable”, and even “ineluctable”.’6

This is in substance the Humean answer, but transposed from psychological into epistemological terms. Hume urges that our idea or pseudo-idea of some necessity in things is really only the misplaced offspring of our felt impressions of necessity; impressions which are in turn the product of the strong habits of association which, as a plain matter of psychological fact, lead us to expect like causes to be followed by like effects. Carr suggests that all talk of inevitability is merely an empty piece of rhetoric which, however misleading, is none the less extremely tempting in cases where we have, or could have had, the very best of grounds for anticipating the occurrence of the event so characterized.


This sort of answer, in the epistemological version favoured by Carr, does contain at least part of the heart of the matter. Where Carr goes wrong in this, again, like Hume is in suggesting that it does, or should, represent also the end of the affair. The first essential and correct point is that to be in a position to say with assurance that something will surely happen is not necessarily to be in a position to say that it is inevitable that it will happen. In the Presidential election of 1956 the conjunction of factors leading anyone who knew his America to expect the return of Mr Eisenhower was indeed overwhelmingly strong. But this particular conjunction of factors provided no ground for suggesting that the polling was going to be conducted fraudulently, or under duress. The plain, perhaps regrettable, fact was that the great majority of the American people still actually preferred Eisenhower to Stevenson.

Again, from the proposition He knows that Isaac will marry Sheila it follows necessarily that Isaac will marry Sheila. But from He knows that Isaac will marry Sheila it does not necessarily follow that Isaac will, necessarily, inevitably, and unavoidably marry Sheila. You may very well be in a position to know that Isaac will marry Sheila without its being the case that poor Isaac cannot help himself. Sheila, after all, may be a very attractive proposition and Isaac a very lucky man. In which case, though there may be no reason to doubt that Isaac if he so wished could renounce the world and flesh and Sheila, you may be perfectly entitled to say that you know that in fact he will do no such silly thing; and that the marriage which has been arranged will duly take place.

These examples, of Eisenhower’s re-election and Isaac’s romance, show how the possibility of knowledge of some future human behaviour does not necessarily preclude the possibility that this same behaviour may be the uncoerced manifestation of free choices. The same examples can be used to illustrate two sources of the popular misconception to the contrary. The Eisenhower case brings out that and how it sometimes possible to forecast the outcome of a free and genuine election. But, of course, the world being as it unfortunaely is, perhaps the more familiar case of ‘electoral’ prediction is different. Here confidence is grounded on the knowledge that the powers that be in some less happier land can as usual be relied on to deploy in their own interests whatever measures of coercion and fraud they consider necessary. The temptation is to mistake it that it is the possibility of foreknowledge as such, rather than the actuality of coercion, which is incompatible with freedom.

The example of Isaac and Sheila can be employed to bring to light in its most simple form a conderably more sophisticated source of the same error. The temptation is illegitimately to transpose the adverb necessarily, and to insert it without any warrant into the conclusion. Properly this adverb should qualify the follows, which links one proposition to another logically derivable from it. In the false conclusion it will qualify whatever verb that proposition may happen to contain. We are thus led to misconstrue a sort of necessity which can only apply to the logical relations between propositions as if we had here a case of some more solid and practical inevitability. The fallacy involved is the same as that in the fatalist argument embodied in the song which begins: ‘Che sara, sara: whatever will be, will be.’ From This will happen it follows necessarily, although unexcitingly, that This will happen. But it certainly does not follo w th at This will happen necessarily, inevitably, and unavoidably.


So far so good. However, as we have seen, Carr goes on to suggest that history could be written, albeit more drably, without using either the term inevitably or a n y o f i t s numerous synonyms and near synonyms: ‘Nothing in history is inevitable except in the formal sense that, for it to have happened otherwise, the antecedent causes would have had to be different.’ This altogether too short a way with the problem. Indeed, Carr’s further suggestion is scarcely even consistent with the argument offered in support of the first and more fundamental contention: to say nothing of its intrinsic implausibility. For whatever would be the point of labouring, as Carr does, to show that from the fact that some actions can be forecast it does not follow that these actions must be in any substantial sense inevitable or unavoidable, if there were not any important, relevant, and substantial senses of these terms? And of course in fact there are.

The words under discussion are usually employed to help in marking the practically vital distinction: between those occurrences which might be (or might have been) prevented, provided suitable steps are now taken (or had then been taken); and those other events which will happen (or would have happened), regardless of any measures which are taken now (or might have been taken then). The appropriate senses of both inevitable or unavoidable and their opposites are as substantial as anyone could wish. Partly for this reason it is entirely consistent: both to recognize that in fact some things are, and others are not, avoidable in this ‘practical’ sense; and to insist that, in what Carr calls ‘the formal sense’, absolutely everything is unavoidable since ‘for it to have happened otherwise, the antecedent causes would have had to be different’.

Perhaps the epithet formal is not quite apt. For this ‘formal’ unavoidability is synonymous with determinism: and that on Carr’s own definition at least looks to be a pretty solid piece of doctrine. But what is surely absolutely clear is that determinism, or unavoidability, in this ‘formal’ sense can  leave plenty of room for the undeniable fact that many occurrences are in the ‘practical’ sense preventable. For there is no inconsistency in saying both of two things: first, that the debacle might have been prevented, that victory had been a practical possibility; and, second, that disaster was nevertheless in another and ‘formal’ sense inevitable. In saying the first thing you would be maintaining only but very importantly that had people taken the appropriate and vigorous steps things would have turned out quite differently. In insisting on the second you would be contending that, the past having been what it had been, it would have been at least in principle possible to have known that the right people would not in fact show the vigour and determination required.


This sort of distinction between the ‘formal’ and the ‘practical’ senses of inevitability is of the very greatest importance, both for the theoretical investigations of history and for the practical deliberations of common life. That it is equally essential to both ought to occasion no surprise. The subject matter of both sorts of inquiry is the same: notwithstanding that typically at any rate practical deliberations are focussed on the future while the preponderant concern of the historian is with the past. Both treat men and affairs. It is dangerously easy, both as an historian and as a man, to confuse the two sorts of inevitability.

You can start by arguing that something was bound to happen; simply on the grounds that the person or people concerned wanted things that way, and there was nothing to stop them. Here, presumably, only ‘formal’ inevitability is involved. Then, overlooking this, you may go on to infer that, since it was bound to happen, it could not have been helped. But if it could not have been helped, by anyone at all including the ostensibly responsible protagonists then there can be no question of holding anyone to account. For, granting that everything was necessary and unavoidable, the situation offers no purchase to the notion of moral responsibility.

However, when once the crucial distinction is grasped, the fallacy in this sort of argument becomes obvious. The second step simply does not follow from the first, because bound to happen is there being interpreted not in the ‘formal’ sense it has in that first move but in the ‘practical’ sense which it has to be given if it is to have the force to generate the eventual and disquieting conclusion. (We thus have an example here of The Fallacy of Equivocation. But there is no point in knowing or using the label except in so far as this helps us to spot and to avoid such false moves.)

There is a further major source of both theoretically and practically important confusions which may be, though it is not necessarily, connected with the fundamental failure to appreciate and to provide for this sort of distinction between two senses of inevitability. You may overlook that, whereas in the ‘formal’ sense inevitability may be very much a matter of point of view of person, time, and place. Seen from Copenhagen in 1940 the German occupation of Denmark was clearly inevitable. But this by no means implies that Hitler and the German High Command really could not help themselves. It does not even imply what is no doubt independently true that there never was anything the Danes could have done to prevent their country being overrun.

‘Practical’ inevitability may thus be relative; though this need not make it any the less an issue of objective fact whether at some specific time some particular occurrence could, or could not, by the efforts of some particular person or collection of people have been prevented.

This essentially relative character of ‘practical’ inevitability is like so much of importance, both in philosophy and elsewhere quite obvious once it has been pointed out. But this does not mean that it must have been impossible for any man of sense to have ignored it. Furthermore, suppose you are inclined to believe that there is only one sort of inevitability, and that the ‘formal’, then you will have no room for the relativity which belongs only but essentially to the ‘practical’ kind.

If, but only if, this point about relativity is taken, it also becomes obvious that to say that some action was ‘practically’ inevitable is no more necessarily inconsistent with saying that it could nevertheless have been helped than to hold that everything is ‘formally’ inevitable is necessarily inconsistent with holding simultaneously that many things are in fact ‘practically’ preventable. In 1956 there was nothing the Hungarians could have done which would have prevented the Kremlin from reimposing a puppet regime behind the screen of Soviet tanks. Seen from Budapest, therefore, this was, you might say (in the ‘practical’ sense) an historical necessity. Again, for anyone knowing the nature of the brutes, it was to be expect that Mr Khruschev and his associates would act as they did. This means that their actions were, in the ‘formal’ sense, inevitable. But neither the first nor the second of these grim facts provides the slightest ground for excusing a deeply characteristic piece of Russian colonialism. For neither constitutes any reason for saying that the leaders in the Kremlin could not help themselves.


Berlin’s main purpose in writing Historical Inevitability was, as Carr mischievously acknowledges, to defend the idea of human responsibility against all and any attacks, whether open or implicit, launched in the name of historical determinism. He is nevertheless prepared to dismiss in a long footnote the saving contention that determinism in general is not necessarily incompatible with the fact of choice, while yet making no serious attempt to come to grips with the one considerable argument which is peculiar to the historical species of the genus.7 That is the argument, urged by Carr and by Hume, that historical determinism is both a presupposition and, to some extent, a finding of historical inquiry. It is an argument which certainly can be strongly challenged. For instance, there seems to be at least one very common and generally acceptable type of historical explanation which tells us not why what happened was at least ‘formally’ inevitable but only how it become possible.8 In so far as explanation of this type can be satisfactory it seems that there may still be room for some measure of indeterminism. Being committed to the opposite view, Carr, again in his own way following Hume, argues instead that though historical determinism is in fact true, the only sort of necessity involved is entirely innocuous.

Yet Carr shows elsewhere in these talks signs that he is nevertheless inclined to accept many of the ideas against which Berlin directed his polemic. This tendency appears most interestingly perhaps in his remarks about moral appraisal in history. These come in the third talk, on ‘History,

Science and Morality’. Carr contends vehemently that to make assessments of this kind is no part of the business of the historian. Now with the milder thesis that moralizing can be so intrusive and so excited, as to get in the way of historical understanding, there can be no dispute. Carr most appositely quotes from an historian of another revolution: ‘It is unfortunate, though very natural, that the history of this period has so generally been written in hysterics. Exaggeration abounds, execration, wailing; and, on the whole, darkness.’ (Carlyle) But the stronger contention, demanding a detachment both systematic and total, is quite another matter.

Discounting a certain amount of rhetoric and ridicule, the only supporting reason explicitly offered by Carr comes as a question: ‘What profit does anyone find today in denouncing the sins of Charlemagne or of Napoleon?’ This is an extraordinary remark; and the more so coming from someone so rightly concerned to insist both on the importance of interpretation in history and on the possibility indeed the inevitability of learning or mislearning lessons from it. The example of Napoleon, too, is quite peculiarly inept to Carr’s purpose. A glance at such a well-known work as Pieter Geyl’s Napoleon: For and Against (Cape, 1949) will serve to remind us of how the debates of historians over the true assessment of that man have been, and seem likely to remain, part of the stuff of political thinking in France.


We have, surely, to look for some deeper reason to explain Carr’s attachment to an ideal of the historian above, or detached from, all moral considerations. Any suggestion is bound to be speculative. Partly no doubt it is a matter of failing fully to assimilate the implications of his own Humean argument. But perhaps it is also connected with his contention ‘that historians and scientists are engaged in different branches of the same study’.9 For the subject matter of history is in fact importantly and relevantly different from that of any of the natural sciences. History is about people, rather than about things. It is for just this reason that it possesses its peculiar human interest. One consequence is that there is room in history for assessments which are in a wider or a narrower sense moral, whereas in the natural sciences there cannot be. Indeed, it is precisely to the extent that the people of the past become through historical study intelligible to us that the possibility arises of reasonable appraisals. Appraisals of some sort it is impossible to avoid, so long as the vocabulary of history continues to be exactly the same as that of ordinary practical deliberation.


1. The Listener, 27/iv/61, p. 731.

2. Ibid., 25/v/61, p. 921.

3. The Listener, 25/v/61, p. 919.

4. See especially the Inquiry Concerning Human Understanding, § VII. For commentary I may refer to my Hume’s Philosophy of Belief, Routledge and Kegan Paul, 1961, especially Ch VII.

5. The Listener, 11/v/61, p. 815.

6. Ibid., 11/v/61, p. 816.

7. Loc. cit., pp. 26-7.

8. W. Dray, Laws and Explanation in History, Oxford University

Press, 1957, Ch VI.

9. The Listener, 4/v/61, p. 773.