Liberty: Incorporating "Four Essays on Liberty"

лM

'A magnificent and indispensable volume: the best introduction to the most important and enduring of Berlin's ideas'

JOHN GRAY

Also by Isaiah Berlin

karl marx the age of enlightenment

Edited by Henry Hardy and Aileen Kelly russian thinkers

Edited by Henry Hardy

concepts and categories against the current personal impressions the crooked timber of humanity the sense of reality the roots of romanticism the power of ideas three critics of the enlightenment freedom and its betrayal

Edited by Henry Hardy and Roger Hausheer the proper study of mankind

LIBERTY

ISAIAH BERLIN

Incorporating Four Essays on Liberty Edited by Henry Hardy

With an essay on Berlin and his critics by Ian Harris

OXFORD

UNIVERSITY PRESS

OXFORD

UNIVERSITY PRESS Great Clarendon Street, Oxford 0x2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship and education by publishing worldwide in Oxford New York

Athens Auckland Bangkok Bogota Bombay Buenos Aires Cape Town Chennai Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kolkata Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paolo Shanghai Singapore Taipei Tokyo Toronto Warsaw

with associated companies in Berlin Ibadan

Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries

Published in the United States by Oxford University Press Inc., New York

Selection of and introduction to Four Essays on Liberty © Isaiah Berlin i 969 'Five Essays on Liberty' © The Isaiah Berlin Literary Trust and Henry Hardy 2002 'Political Ideas in the Twentieth Century' copyright Isaiah Berlin 1950, © Isaiah Berlin 1969 'Historical Inevitability' copyright Isaiah Berlin 1954, © Isaiah Berlin 1969, 1997 'Two Concepts of Liberty' © Isaiah Berlin 1958, 1969, 1997 'John Stuart Mill and the Ends of Life' © Isaiah Berlin 1959, 1969 'From Hope and Fear Set Free' © Isaiah Berlin 1964 'Liberty' © Isaiah Berlin 1995 'The Birth of Greek Individualism' © The Isaiah Berlin LiteraryTrust and Henry Hardy 1998 'Final Retrospect' © The Isaiah Berlin Literary Trust and Henry Hardy 1998 'The Purpose Justifies the Ways' © The Isaiah Berlin Literary Trust 1998 Letter to George Kennan © The Isaiah Berlin Literary Trust 2002 'Notes on Prejudice' © The Isaiah Berlin Literary Trust 2001 This selection © Henry Hardy 2002 Editorial matter © Henry Hardy 1997, 2002 'Berlin and his Critics' © Ian Harris 2002 Illustrations © The Isaiah Berlin Literary Trust 2002 The moral rights of the authors and editor have been asserted

Four Essays on Liberty issued as an Oxford University Press paperback 1969 Liberty published in hardback and paperback 2002

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organisations. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above

You must not circulate this book in any other binding or cover and you must impose this condition on any acquirer

British Library Cataloguing in Publication Data

Data available

Library of Congress Cataloguing in Publication Data

Data available

ISBN 0-19--924988-1 (hbk) ISBN 0-19--924989-X (pbk)

Typeset in Monotype Garamond by Deltatype Ltd, Birkenhead, Merseyside

Printed in Great Britain by T. J. International, Padstow, Cornwall

To the memory of Stephen Spender

1909-1995

The essence of liberty has always lain in the ability to choose as you wish to choose, because you wish so to choose, uncoerced, unbullied, not swallowed up in some vast system; and in the right to resist, to be unpopular, to stand up for your convictions merely because they are your convictions. That is true freedom, and without it there is neither freedom of any kind, nor even the illusion of it.

Isaiah Berlin, Freedom and its Betrayal'

¹ London and Princeton, 2002, pp. 103-4- The lectures that comprise Freedom and its Betrayal were delivered in 1952. (Berlin uses the words 'freedom' and 'liberty' interchangeably.)

CONTENTS

Illustrations viii

The Editor's Tale ix

five essays on liberty

Introduction 3

Political Ideas in the Twentieth Century 55

Historical Inevitability 94

Two Concepts of Liberty 166

John Stuan Mill and the Ends of Life 218

From Hope and Fear Set Free 252

other writings on liberty

Liberty

The Birth of Greek Individualism 287

Final Retrospect 322

autobiographical appendices

The Purpose Justifies the Ways 331

A Letter to George Kennan 336

Notes on Prejudice 345

Berlin and his Critics by Ian Harris 349

Concordance to Four Essays on Liberty 367

Index 371

ILLUSTRATIONS

A page from the proofs of Four Essays on Liberty xvm

The front cover of the first impression of

Four Essays on Liberty xxm

The source of the title 'Five Essays on Liberty' xxxiv

Berlin's notes for 'My Intellectual Path' 282

Berlin aged twelve, Arundel House School, July 1921 329

The first page of the manuscript of

'The Purpose Justifies the Ways' 330

The first page of the typescript of 'Political Ideas in the

Romantic Age' 348

The end of the bibliography from the Isaiah Berlin Virtual

Library, http://berlin.wollf.ox.ac.uk/, October 2001 365

THE EDITOR'S TALE

Liberty is the only true riches. William Hazlitt¹

In the year that Isaiah Berlin died, I was invited by The Times Higher Education Supplement to contribute to their 'Speaking Volumes' series, in which readers write briefly about the book that has influenced them most. I had no hesitation in choosing Berlin's Four Essays on Liberty, which not only bowled me over when I first read it, but also set me on course towards becoming Berlin's editor, and so led, thirty years on, to the publication of this expanded edition of the book.

My THES piece was written just before Berlin's death, and published shortly thereafter.² Part of what I said seems to me to bear repeating in the present context:

I had no idea when I joined Oxford's Wolfson College as a graduate student in 1972 that I was about to discover my eventual occupation. The College's President was Isaiah Berlin. It was clear as soon as I met him (at a scholarship interview for which I arrived late after a car accident, and during which he repeatedly went to the window to see if a taxi had arrived to take him to a lunch appointment) that he was a remarkable man; but I had never read any of his work, and knew next to nothing about him.

I asked where I should start, and was rightly directed to Four Essays on Liberty, published three years earlier. I took it with me on a visit with friends to a remote Exmoor cottage during a University vacation, and was transfixed. Berlin liked to refer to the unmistakable sensation

' From 'Common Places' (1823): vol. 20, p. 122, in The Complete Works of William Hazlitt, ed. P. P. Howe (London and Toronto, 1930-4).

² Issue dated 21 November 1997, p. 21. Berlin died on 5 November. The article is also available on line at http://berlin.wollfox.ac.uk/, under 'Writing about Berlin'. I have slightly adapted the extract used here.

of 'sailing in first-class waters', and this was the sensation I experi­enced. Quite apart from the persuasiveness of the propositions contained in the book, here was obviously a man of rare insight into human nature, a man plentifully endowed with that 'sense of reality' that he welcomed when he found it in others. There was room for disagreement on this or that point, but on the large issues one felt in safe hands.

The central plank in the book is Berlin's value pluralism, his belief that the values humans pursue are not only multiple but sometimes irreconcilable, and that this applies at the level of whole cultures - systems of value - as well as between the values of a particular culture or individual. It is an essential characteristic of the great monistic religions and political ideologies to claim that there is only one way to salvation, one right way to live, one true value-structure. This is the claim which, when it is given fanatical expression, leads to fundamen­talism, persecution and intolerance. Pluralism is a prophylactic against such dangers. It is a source of liberalism and toleration - not just the unstable kind of toleration that waits for the mistaken to see the light, but the deep, lasting toleration that accepts and welcomes visions of life irretrievably different from those we ourselves live by.

Four Essays is full of other gold, including the devastating critique of historicism and determinism in 'Historical Inevitability', the famous discussion of 'positive' and 'negative' freedom in 'Two Concepts of Liberty', and the examination of the tensions in Mill's views in 'John Stuart Mill and the Ends of Life'. It is one of the richest and most humane books I have ever read, and it has deservedly become a classic.

This said, it may seem lese-majeste to tamper with it now, but, as will soon become clear, the first stage of expansion was devoutly wished for by the author himself, and I see myself as taking the process further towards its logical conclusion.

I do not apologise for having put pluralism rather than liberalism centre stage in my comments on Four Essays, though others would invert this priority. Berlin's pluralism seems to me the deeper and more original thesis - which is not to deny the indispensability of his version of liberalism, or of the view of humanity that lies at its heart, a view in which freedom of choice among incommensurably multiple possibilities is central. Indeed pluralism and liberalism, the two leading components of Berlin's philosophical outlook (some­times aptly called 'liberal pluralism'), are mutually interdependent and supportive,[1] and I have at times thought of giving this collection a title such as Freedom and Diversity, but the Occamist imperative, reinforced by the pragmatic desirability of echoing the well-known earlier title, won out.

FIVE ESSAYS ON LIBERTY

The time has come said Linnet to Stallworthy to talk about Berlin again.

Oxford University Press memo from Catherine Linnet, New York, to Jon Stallworthy, London, 21 June 1967

Berlin's oeuvre has been described by Ira Katznelson, somewhat sweepingly but quite understandably, as 'both correct and bold':¹ the luminous, settled, assured qualities of Berlin's writing are widely recognised and appreciated. But there is a paradoxical relationship between these undoubted attributes and the tortuous and tortured route by which his publications came to take the form they do. The 'correctness' is not achieved at the first attempt, nor even at the nineteenth; and the boldness is not matched by an equivalent self-confidence. As Berlin wrote to Karl Popper in gratitude for his approval of Two Concepts of Liberty, 'I have little confidence in the validity of my own intellectual processes.'² Although he commanded the stage, he trembled in the wings.

The genesis of Four Essays on Liberrf was just as chaotic and prolonged as that of the other compilation of his essays that Berlin published before I became his editor, namely Vico and Herder The Oxford University Press file on the book is a treasure-house of anecdote: frustration, misunderstanding, tergiversation, indeci- siveness, prevarication, unrealistic expectations abound. The whole

excellent Isaiah Berlin (London, 1995), who believes that Berlin's pluralism narrows the field for the justification of his liberalism: see Gray's chapter 6, 'Agonistic Liberalism'.

| 'Why is it so intuitively true that Berlin's work is both correct and bold?' he asks in 'Isaiah Berlin's Modernity': Arien Mack (ed.), Liberty and Pluralism [Social Research 66 No 4 (Winter 1999)], 1079-101, at 1079.

Letter of 16 March 1959.

Published by OUP in London and New York in 1969. Bibliographies often state, misleadingly, that the book was published in Oxford.

I offer a brief version of the saga of this later (1976) volume in Berlin's Three Critics of the Enlightenment: Vico, Hamann, Herder (London, 2000: Chatto and Windus; Princeton, 2000: Princeton University Press), pp. vii-viii.

proceedings, year after year, are accompanied by frantic re­schedulings on the part of OUP, as well as complementary and conflicting discussions of other projects, which appear out of the fog and then recede. OUP become increasingly desperate as time slips by, and some of the wry internal memoranda make excellent reading. I say all this not to poke fun, though the file is undoubtedly fun to read, but because we learn much about Berlin the man by having the complex process of creation of his famous and important book - in his view, his most important book - laid bare in such comprehensive detail. I hope it is clear, too, from my opening remarks that the spirit in which I tell the story of the book's gradual emergence is one of affection rather than censure, for all that Berlin's conduct, benign but gloriously unprofessio­nal,[2] caused justifiable exasperation on the part of his publisher. The path was stony, but the destination fully worth the journey, and not to be reached by a more direct route.

Here I can only skim off the cream of the story. The file opens in November 1953 with a letter from the New York office of Berlin's literary agent, then as now Curtis Brown, to Oxford University Press, Inc., New York, who had taken the lead in the commission­ing of the book. At this point only the first two of the four essays had been written, though a book of essays 'on political topics' was already under discussion. 'I will try to obtain a list of essays from Mr Berlin as quickly as I can,' writes John Cushman of Curtis Brown. What would he have said, we may speculate, had he known that it would be sixteen years before the book finally appeared?

At that time OUP had two publishing offices in the UK, one in Oxford (the academic Clarendon Press) and one in London, at Amen House. Amen House was responsible for publications aimed at a general readership, including Oxford Paperbacks, the series in which the UK edition of the book was to appear. The London Publisher, Geoffrey Cumberledge, was interested but pessimistic: 'Berlin . . . is brilliant but his output is very small and his performance is worse than his promise.'

In I 9 58 Berlin gave his celebrated inaugural lecture as Chichele Professor of Social and Political Theory in Oxford, 'Two Concepts of Liberty', and in 1959 his Robert Waley Cohen Memorial Lecture, 'John Stuart Mill and the Ends of Life'. Both of these thereafter start to appear as constituents of the volume, by 1960 hyperbolically if provisionally entitled 'Collected Writings' by the New York office.

In reply to an enquiry from New York early that year about progress, Colin Roberts, Secretary (that is, head) of OUP, writing from the Clarendon Press, quotes a letter from Berlin, the first communication from him represented in the file:

Alas, my Introduction to the paperback on liberty is not just a question of a willing typist - I wish it were - last-minute corrections are my metier as you know too well, but it is not that that is delaying me. I should like to write a preface - more a postface - in the way of discussing and, so far as I can, replying to the various points and objections which all three essays¹ have encountered one way and another - not indeed by name and address, but in fairly general terms. This I cannot do for a while - I am a slow worker - and hope to do in summer.

In March I 961 Amen House writes to OUP's Deputy Secretary, Dan Davin, at the Clarendon Press: 'Is there the vaguest possible chance that Berlin might even have begun to work on the prefaces which he insists are necessary?' A letter from Berlin reported by Davin later that month announces that

The Three Essays have now become four - Mill being added ... As to the Introduction, I shall write it in the summer in July and August, it will have to be in the nature of a general reply to all the many and fierce objections that have been made to these essays, and are still being made in current publications, so that the Press in New York must not think they are losing something with every new reference in my reply to the critics. They will acquire at least one new potential reader (the latest onslaught is in a magazine called Dissent, which arrived yesterday)² - so long as my opinions to my own astonish­ment provide a live horse for the critics to flog, it will not be too late to re-issue the essays.

Answering an enquiry from John Brown (Cumberledge's suc­cessor), Berlin's typist Olive Sheldon writes on his behalf in September that he is at work on the Introduction to a book to be

' The essay on Mill had not yet been added. At this stage the work is usually referred to as 'Three Essays on Liberty'.

^г David Spitz, 'The Nature and Limits of Freedom', Dissent 8 (1961-2), 78-86.

called 'Essays on Liberty' or 'Against the Current' or 'Against the Stream'. Through her he expresses doubts about the value of the essays on J. S. Mill and on twentieth-century political ideas and suggests that they be sent to a referee. The Introduction is promised for January 1962. In November Harold Beaver of Amen House writes to Catherine Linnet in New York: 'I feel sure that Berlin is merely flapping when he wishes his material to be read.' Read it was, however, by Adam Ulam, Professor of Government at Harvard, who reported favourably, as expected, prefacing his remarks with this sound observation: 'I am not entirely in sympathy with the custom of sending the work of a reputable scholar which has a style and point of view of its own to be picked and hacked at by somebody else.'

In January 1962 Berlin writes a letter to John Brown that is worth quoting in full:

I am oppressed by feelings of guilt about the Introduction to the paperback containing my various essays on liberty and generally related topics. I do not believe I shall achieve this Introduction before the Summer. The reasons for this are: (1) that since it involves reading the accumulated criticisms of the various ingredients of this volume - that was the point of the new Introduction - [it] needs a good deal of time and deliberation and careful drafting of answers to objections. Critical reviews seem never to cease although I am prepared to draw a line at i January 1962 and take into consideration nothing that appears thereafter.

(2) Living the life that I do, I deliver too many lectures outside my Oxford curriculum, sit on too many committees, and generally scatter such energies as I possess in a highly uneconomic and indeed often absurd manner. In my lucid moments I regret this very much and make constant resolutions to resist invitations by undergraduate societies, and to lead a rational, i.e. more concentrated, life. But all these excellent resolutions break against the barrier, and the feeling that as a Professor I cannot refuse to tell the truth to those who make quite a good show of appearing to want to hear it. As for the committees, since they are my only excuse for going to London or abroad, I secretly cling to them even though I recognise their time- eating and energy-destroying properties.

These things being so, I know myself well enough to realise that I cannot write this Introduction in term-time - in April I shall be away both lecturing and functioning on my committees - but I shall write my piece in May or June, and you shall have it by mid-July. I felt it to be only fair to you to let you know how the matter stands - if this delays publication, then, so far as I am concerned, I shall shed no tears, but I sincerely hope that it will not interfere with your publishing plans too much.

This generates a note from Beaver to Linnet: 'Isaiah Berlin, the great cunctator, has again put off supplying the preface.'

In May Bud MacLennan of Curtis Brown asks John Brown for an advance of Јioo, and in his absence a colleague tells her that they can pay £50 or £75, 'but I do not think we can go beyond this figure'. (One wonders what OUP's estimate was of the likely sales of the book, which has remained in print and in constant demand ever since.) The contract for what was now to be called Four Essays on Liberty was signed in July, replacing an earlier contract of July 1959 with New York for Three Essays. In October John Brown writes to Sheldon Meyer in New York: 'I think we have got everything satisfactorily tied up, provided only that Berlin will produce the copy.'

Berlin writes to John Brown in February 1963 that 'the Intro­duction for Four Essays on Liberty is a ... complicated matter', partly because he was giving priority to another project (which, like many others, did not materialise), a book based on the 1962 Storrs Lectures at Yale, 'Three Turning-Points in the History of Political Thought'.

In March 1964 Jon Stallworthy of Amen House, by then in charge of Oxford Paperbacks, writes to Curtis Brown that 'it is over a year since we last corresponded about the Introduction for Sir Isaiah Berlin's Four Essays on Liberty and I wonder whether you could give us any news of progress on this?' The reply is that the piece will not be ready for at least another year, and OUP are asked if they wish to cancel the contract. Stallworthy writes to Peter Sutcliffe in Oxford: 'The Preface has been promised us for the best part of four years, and I think everyone - including perhaps even Berlin - realises that we shall never see it now.' Stallworthy asks Curtis Brown for permission to go ahead without it. Richard Simon of Curtis Brown replies that Berlin will defi­nitely produce the Introduction for April i966, and that, if he doesn't, OUP may publish without it. This arrangement is accep­ted by Stallworthy.

Needless to say, this deadline slipped, ostensibly because Berlin was ill. Stallworthy secured permission to typeset the four essays before the arrival of the Introduction.' Before sending the typescript to the printer, he consulted Berlin about two possible forms of typesetting - hot metal and Monophoto - and explained that, if there were to be changes, it was vital to opt for hot metal. Berlin undertook to make no changes, and Stallworthy, rashly believing him, opted for Monophoto.² The Introduction was re- promised for the end of August, again on the understanding that the book would appear without it if it were not ready in time.

A further reversal occurred when Berlin wrote in the following terms to Stallworthy four days short of the new deadline, in a letter signed on his behalf in his absence by his secretary, Baillie Knapheis:

[. . .] I should like to hasten, in the first place, to thank you for your extremely considerate and patient treatment of me - beyond my deserts. I know that the Oxford Press in New York must regard me as a highly unsatisfactory client - because of all these delays - but one of the secret causes of this is my suspicion that the works which they kindly wish to reprint as a paperback are in some cases scarcely worth it; I have looked through 'Historical Inevitability' again, and I find that there are all kinds of things wrong with it, and I should certainly be ashamed if it appeared in an unaltered form. I have gone through the disagreeable task of reading through the nastier criticisms of it - such as I have kept - the more violent and ephemeral I mislaid or lost almost at once - and it appears to me that what some of the critics said is true, and that, in the interests of the readers and general integrity, the text cannot be left wholly intact. Consequently I have introduced corrections - though far less radical ones than were perhaps required - and hope to make up for this in the Introduction, which I propose to prepare next week. In the meanwhile I do hope that the corrections will not reduce the Press to despair: I realise that there is something for the printers to do/ and if this is regarded as financially awkward, I am so anxious for this labour to be done - that is, for the corrections to be introduced (I should be ashamed - and indeed could not conceive the prospect - of letting the texts go out unaltered), that I should be prepared to consider reimbursing the Press for these

| This is why roman numerals are used to paginate the Introduction in Four Essays.

It seems he had not studied the file for 'Two Concepts of Liberty', where, with impressive self-restraint, Colin Roberts writes to Berlin on 6 November 1958: 'You have certainly had a field day with the proofs.' The lecture had to be completely reset.

One of the great understatements of our time: in the end the whole book had to be reset.

unexpected expenses. In fact the only prospect I could not contemplate was for the corrections not to be incorporated.

I hope you will forgive me for being such a nuisance. I know all authors are, and am perhaps not the worst among them; nevertheless, unlike some authors, I do possess a genuine conscience with regard to publishers and do not regard them as mere philistine adversaries to be sparred with, but as genuine intellectual collaborators, particularly the Press. Consequently I do hope that you will once again be patient with me, again beyond my proper deserts - for I am quite clear that if the only condition for publication is that the texts should go out unaltered, I would rather nothing were published at all, and that these essays continued to dwell in their present decent obscurity [.. .]

Mercy, rather than justice, is, I suppose, what I am asking for: but I truly cannot see how you could deny it to me. You must have had authors far more tiresome than even myself. Perhaps what I am asking for is not so terribly unreasonable. At any rate, I am very grateful.

Page proofs of the four essays arrived at the end of November, but there was still no Introduction. This finally arrived in May 1967, but was immediately put on hold because Berlin wanted comments from Stuart Hampshire and Herbert Hart. In the meantime he continued to correct the essays themselves heavily, despite his promise not to do so. This elicited the following comment from between Stallworthy's gritted teeth:

I think I should mention [a tactful substitution for 'remind you'] that the book has been set up by a Monophoto machine that produces a page not of lead but of film negative. Every correction involves a delicate operation not unlike that for the removal of a cataract from a human eye; the skin of the negative has to be cut and a new line or letter grafted on. Such corrections are very expensive.

Berlin finally returned the corrected proofs of the four essays in August. A month later he sent OUP a revised text of his Introduction, writing in his covering note: 'Owing to the devastat­ing criticisms it has received, I have altered it, not nearly as much as the critics wish, but still, perhaps sufficiently to avoid howling errors (or perhaps not).' At this point an internal OUP note from Stallworthy reads: 'Despite all my explanations about the cost of correcting a filmset text, my suggestions, pleas, further explana­tions, further suggestions, and further pleas, Berlin has made extensive corrections.' If only the book had been published in the days of word processors and modern typesetting technology.

Ш --.-ХГ С Г" ' Ъ Ч

rm R J;. hsns <

were established as an accepted . truth,,our world-would be ■ *f\\ns*

formcd^^more radically rhjn iyns the ideological world of the

dassical and middle ages by tin- triumphs of mechanistic prin- tiplcs. or those of natural - selection. Our words our modes of s|*.ech and thought — would be transformed in literally un­imaginable ways; the no9<>ns.of . choice, . - . a^^^^if resprnsibihty,. freedom, Ire. ««'deeply emledded ..in. our outlook, thai our in* rift as , te^i nils in .1 world genuinely lading in these concepts, can, l slutij)u maintain, be conceived by us . only with the greatest d'riu'ij' \VHin there is, as yet, no need to alarm our­selves undul». We arc . speaking.only .of pscUid«^.icntihc ideals; the reality is . nut in sight:. The evidence for a thoroughgoing determinism is not to hand; and if there is a persistent tendency" to believe in it in some theoretical fashion, Ш.i! is «m-l^duf lai more to the lure of a 'scientistic' or mttaphy чшагымп or i'«at, on . the part ofth01>e who dc-sire to change society to believe tile stars in their courses are fighting for theni/<>r.it'jx." .\i*() fy'zrx to a longing to bay down moral burdens, or minimize milt, uHul ^ • responsibility and transfer it to impersonal ibi ces which can be ■aj^fi accused of causing all o«r discontenis^ilian to any iiKTeasI^f our powers nt i n'tjwl Tfftmiw nr;TTi"i*impriivrnirtn in our 5 Btific techniques. Belief in historical delemimism of this гК pe is, of Course, very widespread, particularly in what 1 should У} 1 (j, i > i tike to call its 'ЬшогппорМЫ' form, by which 1 roeati roeta- r*j ^ physico~theological theories of hisiorv, which attract many who £•4. V have lost their faith in older religious on ho false fear in attribtiting respotwi- rf bffitv to individuals Their specific attributions apd Mi^ksft ntav "p he tttMtfay hut both they and their readers would

Ч Г ИШ - «

^J 1 I

■r-^jjj^' Л

( l.J,-'

К К TV

5 ⁴ К

Pi

^:

uik

| Jx: ^ ":(.,!i their \vr\ activity had been superseded and stultified by the advances of sociology. or bv some deeper meta- phwuil msjght, tike that of oriental stargtaers by the dis-

In November Stallworthy sent Berlin a long list of queries about the final text of the Introduction, but it was February 1968 before Berlin replied. In his letter (reproduced on page 2 below) he wrote:

I see that gradually but inexorably I am becoming if not your most intolerable (though I may be that too) certainly your most time- consuming author. At the risk of inflicting a blow upon you which may seriously endanger your health - such health and optimism as you may have regained during your recent holiday -1 propose to inflict yet another hideous blow upon you [...] It has been represented to me by kind friends (for once genuinely kind) that the book might be improved by the inclusion in it of yet another essay on the same subject, namely my Presidential Address to the Aristotelian Society a few years ago, the title of which was 'From Hope and Fear Set Free'. This would make a fifth essay in the book and the title could be altered from 'Four Essays on Liberty' either to 'Five Essays on Liberty' or simply 'Essays on Liberty', since five essays perhaps begin to deserve that title. The piece in question is not the worst that I have written, and I should like it included.

He enclosed the necessary small changes to the first paragraph of the Introduction, and added in a covering manuscript note: 'I do indeed grovel before you: I cannot operate any differently from the way that I do: but why should you (or the printer) suffer? Determinism & the helplessness of man must be true after all.' Stallworthy's reply on the fifth essay was this:

Tempted as we are by the thought of a fifth essay, I'm very much afraid that it is now too late to include this. We have advertised 'Four Essays' in numerous catalogues, have made a block for the cover, have

(toposite) A page from the proofs of Four Essays on Liberty, see pp. 161-2 below. Berlin's long correction, which was not incorporated into the finished book in this form, reads as follows: 'Some thinkers seem to feel no intellectual discomfort in interpreting such concepts as responsibility, culpability, etc. in conformity with strict determinism. I must own that while the notion of uncaused choice, which is nevertheless not something out of the blue, is one of which I know of no adequate analysis, its opposite, a choice fully attributable to antecedent causes mental or physical, and yet regarded as entailing responsibility and therefore subject to moral praise or blame, seems to me even less intelligible. This difference, which has so deeply divided opinion, is the crux of the matter: a puzzle which has exercised some thinkers for more than two thousand years: while others either fail to see it, or have regarded it as a mere confusion. The present state of controversy seems to me much the same as in the days of the Greeks who first began it.'

worked out a published price on the basis of the present length, and - not last and not least - have set up as headline on every other page 'Four Essays on Liberty'.

Berlin replied:

I am naturally disappointed that you should consider it too late to include 'From Hope and Fear Set Free'. I am afraid that no further collection of essays on philosophical topics by myself will ever materialise [. ..] But this essay belongs as of right to the original collection which you are about to publish and, if not included there, can never be reprinted at all. This may seem to you (and, on reflection, to myself) not to be an appreciable loss to anyone; nevertheless, I should like to make a final plea, and beg you to consider whether perhaps it could not be substituted at the last moment for 'Political Ideas in the Twentieth Century', to which it is vastly superior. The changes required will, after all, not be very grave. It will mean the loss of one appendix[3] and one, by now, ancient piece - that really could reappear, if it were thought worthy, in some other collection. I do not underestimate the trouble to which I am putting you, but, for once, my desire to improve the volume - as this substitution undoubtedly would do - is much stronger than even my easily disturbed guilt about all this tiresome chopping and changing for which I have been responsible. Would you give the matter another thought? Could you attempt to soften the (by now) savage breast of your New York colleagues? I do beg you to consider this once again.

Far from softening a savage New York breast, this hardened an Oxford heart. Deciding that the time had come for straight speaking, Stallworthy asked Berlin to come and see him. He now takes up the story in his own words:

Berlin countered with an invitation to lunch in All Souls. 'Thank you, but no,' I replied. There had to be a show-down and I wanted the territorial advantage of my own corral. Berlin, recognising the strat­egy, proposed other meeting-places, pleaded pressure of work, but I said No: there would be no further progress on the book until we had met - at the Press - to discuss the situation. He prevaricated for some weeks, but finally agreed.

I waited for him that morning wearing my darkest suit, my darkest frown.

'Sir Isaiah . . .'

^He interrupted my frontal attack with a raised hand and a rapid

' This now appears as note r to p. 69 below.

diversionary manoeuvre: 'They tell me you're translating Blok. ' Greatest poet of the Revolution. Did you know his wife? No? I met her. Must tell you about her.' And he did - brilliantly.

'Sir Isaiah . . .'

Again the raised hand - and now the diversionary manoeuvre cunningly changed course: 'I know I've been tiresome, but I've been so busy, so distracted by this new College for homeless lecturers.' Thirty-four years later, as a Fellow of that College, I am amused to remember the old magician's revolutionary peroration: 'I will take them from the highways and byways. They will be the sweepings of the streets, but they will inherit the earth!'

He was irresistible. I stuck to my guns over the fifth essay, but weakly agreed to have reset - at OUP expense - the four he had so outrageously revised.

The fifth essay was not, however, banished from the book for all time. In his next letter Stallworthy wrote: 'We are agreed that when a new edition is called for we will add "From Hope and Fear Set Free".' The Stallworthy Treaty of 1968 is being honoured in 2002.

Berlin, by his own admission, over-corrected the proofs of the Introduction 'as usual'. He asked Stuart Hampshire to write a footnote answering the criticisms of his views.[4] He observes to Stallworthy that E. H. Carr would be happy to do the same, 'my God! if let. But the whole piece must not consist of attributions of views (mainly my own) furiously disowned by their putative holders.' Stallworthy replies, having suggested a reduction of the corrections: 'I think it is no exaggeration to say that the present corrections would require the resetting of nearly half the Introduc­tion.' (In the end the whole of it was reset.) In his reply to Stallworthy's pleas, Berlin says that he has endeavoured to make changes that occupy the same space as what they replace. He adds: 'So now we can go - I should like to say full steam ahead, except that I feel that I have held the engine up so long, I cannot complain if it seizes up or moves backwards.' A later letter, answering final queries on the proofs, concludes: 'My doctrines are attacked so ferociously in this year's B.Phil. examination in Politics that I anticipate storms, not from embattled students only, but from every possible quarter, when my unpopular doctrines are pub­lished: that or chilly silence, broken by a few mildly contemptuous dismissals in the TLS and the like. To all this I am resigned, or at least suppose myself to be.'

From now on it is more or less downhill all the way, though there is still a series of minor hitches. In September Stallworthy tells Linnet: 'Berlin continues to fight a harassing rearguard action, but we shall overcome.' The following month a memo from Linnet ventures: 'We are toying with the idea of listing this book in the next seasonal catalog.' When Berlin saw the final proofs in October, supplied only so that he could answer some questions about page references in the index, he noticed that there were still a number of errors in the text; the survival of some of these in the finished book is an additional minor justification for a new edition.

An advance copy was eventually sent to Berlin in March 1969, together with the information that the publication date would be 1 5 May. As had been intended from the start, the book was published only in paperback, as part of the Oxford Paperbacks series. This strategy, in my view (perhaps aided by hindsight), was a mistake, at best a premature publishing experiment, since it played its part in ensuring the noticeably meagre review coverage the book received: the established custom of literary editors, visible even to this day, was to take hardbacks more seriously than original paperbacks.¹ The book may have improved the profile of Oxford Paperbacks, but its manner of publication, possibly rein­forced by its somewhat self-effacing title, damaged its early fortunes.

Berlin's reaction to the advance copy includes the following:

I was naturally horrified to see my own likeness upon the cover - I

had not been warned about this and it set me back a good deal. Is this

' In New York, however, a hardback edition was published in 1970. In 1979, too, when I was myself an editor at OUP, I bound up part of a reprint of the Oxford Paperback in hard covers in an attempted rearguard action, but because of the low-quality paper used for the series at that time this was an unsatisfactory hybrid. Only now is the book being given the kind of physical incarnation it has always deserved.

\

1WSK

■ HLZT^p

ШШШШШШШШВШШ

' ' ' ■: Л

iiBiSI

w Щ ЯГ ЩШ ЩШ ЩВ щШШ

1Ш1Ш1111 waw

НКиИ

Ч'-ЖЖ Ж''"'-::';'

щ^нпн янннрнияи

Яв ■ j

I ■ J

дЦИмдИ

,■::■■■ .-с,.,.. s

1 < Ш fj Щ '

ypf 1

, ,,

; ■

.

■НННИнш. #

■

Нк #ЧИ|

ш ди

,

. •' .

гл

В .

S :

ёыыямяк

ИНИН

ШШт

ЙИ 1 ''

УЛЛМгЛ Шя ■йЦншШш

ТШВШВШ

щЯВШЯгяЯЯЩВЯЯ^ЯЩШ

'ШШШЙшк

Wk>

^ШРШЯИИ

■■■ИМ

ц^нНк, ^шшшШШШИШШШ

Ж

• ' • ' ..■■■■■■ ' -v.,'., ■.■■.. ....

^ V • ^ ^ ■;. ■.-.'.•■-у

... . ,. : . : ..... ..'.

... . ' .. ... ' ■

ЩВрШрв

The front cover of the first UK impression of Essays on Liberty (1969)

absolutely indispensable? However, it is done and I must not cry over what seems to me a slight lapse in taste (do you not agree? secretly). As for the rest, the book looks very nicely done. Now I expect terrible brickbats, though it seems to me about the worst moment for preaching the sentiments for which I do not feel ashamed and which I do not wish to withdraw, but which are regarded by young and old as singularly 'irrelevant' to their preoccupations. However, never mind, perhaps posterity will be kinder or perhaps there will be no posterity to have to be kind. Perhaps it will all be justifiably forgotten - book, author, reactions and all.

He also provided lists of people, nearly 200 in all, to whom he wished copies to be sent at his expense, commenting: 'I expect these are about the only persons who will in fact wish to buy the book - however, never mind.'

Stallworthy replies: 'I was sorry to learn that you are now not happy with the cover. You will remember, I am sure, that Carol Buckroyd called at your house one Sunday morning with a proof. You did not then like the yellow lettering and chose from the books on your shelves a light blue to replace it.' And in its light blue livery, bound with rapidly crumbling glue, the book now finally entered the public domain.

For 'Five Essays on Liberty' - the second edition of Four Essays with which this new collection begins - I have added, for reasons that will already be apparent, 'From Hope and Fear Set Free', finally removing the quotation marks that signalled the Swinbur- nian origin of its title,1 since the accurate but perhaps pedantic punctuation ' "From Hope and Fear Set Free" ' (just as in the case of ' "The Purpose Justifies the Ways"') has seemed to cause more difficulties than enlightenment. I have also edited the text of the original four essays and their Introduction, breaking up some long sentences and paragraphs in line with wishes Berlin had expressed too late to OUP, adding and correcting references, quotations and translations as necessary, reinstating a handful of late alterations overruled by OUP for the first edition on practical grounds, and generally ironing out wrinkles - without, of course, making any

' A line from Swinburne's The Garden of Proserpine.

alterations of substance.[5] 'Two Concepts of Liberty' and 'Histori­cal Inevitability' had already received most of this treatment for their inclusion in the one-volume selection from Berlin's writings published in 1997 as The Proper Study of Mankind, and have not been significantly further revised here. But because Four Essays has been so widely cited in the literature, I have provided a concord­ance showing where the page numbers of the first edition began, so that references to that edition can easily be looked up in this one.

OTHER WRITINGS ON LIBERTY

The reprinting of already published articles is in principle to be reprobated, but in this case there are extenuating circumstances.

A. H. M. Jones, Athenian Democracy (Oxford, i960), p. v

I have also added a number of other writings that bear on the same subject, so that they can all be conveniently consulted together in one place. Indeed, the essay on the Greeks has not hitherto been collected, and the penultimate appendix not previously published. The inclusion of 'Liberty' and of the excerpts from 'My Intellectual Path' entitled 'Final Retrospect' breaches my general rule that the same material should not appear in more than one collection edited by myself:[6] but as these are short items the duplication is perhaps venial, and they do so evidently belong here. 'From Hope and Fear Set Free' is another such exception, of course, since it has already appeared in Concepts and Categories (1978); but the special reasons that apply here have already been made clear. I have wavered about also adding 'Herzen and Bakunin on Individual Liberty' from Russian Thinkers (1978), since it does throw a good deal of light on the topic of the present volume; but it is another full-length piece, and since its approach is more prosopographical, its inclusion here seemed in the end not essential. The other obvious candidate would have been Berlin's 1952 lecture series, Freedom and its Betrayal, this, however, is being published by Chatto and Windus and by Princeton University Press as a separate volume at the same time as Liberty.

Remarks on each of the additional items now follow seriatim.

Liberty

This short summary of Berlin's views on liberty provides a useful orienteering guide for the newcomer. Berlin drafted it in prepara­tion for his appearance in 1962 in an Associated Television film on freedom of speech, the first of a series of five (sic) programmes collectively entitled The Four Freedoms, presented by Bamber Gascoigne. What Berlin actually said in the film is very different from the remarks he prepared in advance, as usually happened; and out of nearly ten minutes of recorded material (a transcript survives) only two minutes were used in the broadcast.

In 1993 Ted Honderich invited Berlin to contribute an article on liberty to a volume he was editing, The Oxford Companion to Philosophy. Berlin did not feel able to write a new piece. He had written nothing substantial since 1988, when he published his intellectual credo, 'On the Pursuit of the Ideal', a response to the award of the first Agnelli Prize for his contribution to ethics.[7]Although his intellect was undiminished, and he continued to compose short occasional pieces, it seemed clear that - reasonably enough in his eighties - he had in effect laid down his authorial pen.

He asked me, however, whether there was anything among his papers that could be made use of; I offered him this short item, which he had dismissed as nugatory when I first drew it to his attention. Slightly to my surprise, therefore, he now found it 'not bad', revised it, and offered it to Honderich, who happily accepted it as it stood.

The Birth of Greek Individualism

It was also in 1993 that Jeffrey Perl, the editor of Common Knowledge, told Berlin, in a letter inviting him to contribute an article, that the journal had been set up under the influence of his work, especially on the subject of pluralism. In his reply Berlin ventured a degree of scepticism about this assertion, but allowed that he felt 'profoundly flattered by the possibility, let alone the probability', of its truth. He also regretfully declined the invitation to write for the journal, partly because of his general disinclination

mentioned above - to undertake new writing, but also because he did not believe he was equipped to deal with the specific topic suggested by Professor Perl.

Not long after Berlin's death I came across this exchange of letters among his papers, and told Perl that, in the light of his original invitation, Berlin's Literary Trustees would be happy to offer him one of Berlin's unpublished pieces. I selected this particular essay because it deals with a topic not covered except in passing in any of Berlin's other publications, and because Berlin himself had told me that he thought something might one day be made of it.

The essay is an edited version of the text Berlin prepared as a basis for the first of his three Storrs Lectures at Yale in i 962; as mentioned above, these were entitled 'Three Turning-Points in the History of Political Thought'. The second and third turning-points

Machiavelli and romanticism - are well covered in his other published essays, especially 'The Originality of Machiavelli', reprinted in both Against the Current (1979) and The Proper Study of Mankind, and 'The Romantic Revolution', which appears in The Sense of Reality (1996). There is also now, of course, The Roots of Romanticism (1999).

Final Retrospect

The two excerpts included under this heading are taken from 'My Intellectual Path', a retrospective autobiographical survey written towards the end of Berlin's life. In February 1996, in his eighty- seventh year, he received a letter from Ouyang Kang, Professor of Philosophy at Wuhan University in China, inviting him to provide a summary of his ideas for translation into Chinese and inclusion in a volume designed to introduce philosophers and students of philosophy in China to contemporary Anglo-American philoso­phy, hitherto largely unavailable to them in their own language.

Despite his de facto authorial retirement, the Chinese project caught his imagination; he regarded this new readership as impor­tant, and felt an obligation to address it. He told the Professor that he would try to write something. With a single sheet of notes before him, he dictated a first draft on to cassette. The transcript was at times rough-hewn, and stood in need of the editing he invited, but scarcely any intellectual additives were needed to produce a readable text. When he had approved my revised version, making a few final insertions and adjustments, he said, with his characteristic distaste for revisiting his work, that he did not wish to see the piece again. It was to be the last essay he wrote. It was published in the New York Review of Books in the year after his death, and also in The Power of Ideas (2000). I have included the two most directly relevant sections here because they bring up to date, albeit more briefly, the view of his critics which occupies much of the Introduction to 'Five Essays on Liberty'. It would have been possible to add other sections, especially those on monism, pluralism and the pursuit of the ideal, but it seemed best to mirror the structure of that Introduction more narrowly.

AUTOBIOGRAPHICAL APPENDICES All central beliefs on human matters spring from a personal predicament.

Berlin to Jean Floud, 5 July 1968

The Purpose Justifies the Ways

Berlin first came to England, as an immigrant, in early 1921, aged eleven, with virtually no English. This story (untitled in the manuscript), which, he told me, won 'a hamper of tuck' in a children's magazine competition, was written in February 1922, when he was twelve.¹ As far as is known, it is his earliest surviving piece of writing, as well as his only story, and shows how far his

' The story is written on headed notepaper from the Royal Palace Hotel, Kensington, where the Berlins stayed while waiting to move to a new address; the sheets have been sewn together, presumably by Berlin's mother. At the top of the first page there is an inscription in another hand, apparently the author's own at a later date: '1. Berlin. February 1922. (author being 12I years of age)'. At the end of the manuscript appears the signature 'I Berlyn'. The Harmsworth weekly magazine The Boy's Herald ran a 'Tuck Hamper Competition' for 'storyettes' at the time, but Berlin is not listed among the prize-winners in early 1922, frustratingly. However, the 'storyettes' are merely humorous anecdotes of around a hundred words: perhaps Berlin was awarded an ex gratia hamper for an impressive contribution in the wrong genre.

English had developed after just a year, as well as his general precociousness.[8]

It is a fictional story about a real person, Moise Solomonovich Uritsky, Commissar for Internal Affairs in the Northern Region Commune of Soviet Russia, and Chairman of the Petrograd Cheka, who was in fact murdered by a Socialist Revolutionary named Leonid Kannegiesser on 31 August 1918. I chose Uritsky's 'motto' as the title because the story so clearly points forward to Berlin's repeated later insistence that present suffering cannot be justified as a route to some imaginary future state of bliss. In this sense the story is the first recorded step on his intellectual journey through life, a journey summarised in 'My Intellectual Path', written seventy-four years later.

Berlin always ascribed his lifelong horror of violence, especially when ideologically inspired, to an episode he witnessed at the age of seven during the February Revolution in Petrograd in 1917: while out walking he watched a policeman loyal to the Tsar, white- faced with terror, being dragged off by a lynch mob to his death. This story surely vividly reflects the power of this early experience, and reveals one of the deepest sources of his mature liberalism.

Letter to George Kennan

Berlin's papers include a mass of often detailed correspondence about the contents of Four Essays on Liberty, both before and after its constituent essays were collected in that volume. Much of this material will in due course be published in its proper chronological place among Berlin's other letters, but there is one letter in particular that stands out from the rest as a powerful statement of the personal vision that lies behind Berlin's work in this area. Berlin liked to allude[9] to a passage in Bertrand Russell's History of Western Philosophy where Russell says that, if we are to under­stand a philosopher's views, we must 'apprehend their imaginative background':

Every philosopher, in addition to the formal system which he offers to the world, has another, much simpler, of which he may be quite unaware. If he is aware of it, he probably realises that it won't quite do; he therefore conceals it, and sets forth something more sophisti­cated, which he believes because it is like his crude system, but which he asks others to accept because he thinks he has made it such as cannot be disproved. The sophistication comes in by way of refutation of refutations, but this alone will never give a positive result: it shows, at best, that a theory may be true, not that it must be. The positive result, however little the philosopher may realise it, is due to his imaginative preconceptions, or what Santayana calls 'animal faith'.[10]

One might discuss the extent to which this picture fits Berlin's own case: for example, Berlin was certainly not unaware of his own 'imaginative preconceptions'. At all events, the letter to Kennan vividly expresses the character of one of the main rooms in Berlin's own 'inner citadel', to use his own metaphor.[11] For this reason I decided to include this letter here in advance of its publication as part of Berlin's correspondence. It was written in response to a warmly appreciative letter from George Kennan about 'Political Ideas in the Twentieth Century', and surely speaks for itself.

Notes on Prejudice

Another room in the citadel is brought to life equally vividly, if more briefly, in some hurried notes Berlin wrote for a friend (who prefers not to be identified) in i 98 1. His friend was due to give a lecture, and wrote to Berlin to ask for suggestions as to how he might treat his theme. Berlin had to go abroad early the day after he received the request, and wrote the notes quickly, in his own hand, without time for revision or expansion. The result is somewhat breathless and telegraphic, no doubt, but it conveys with great immediacy Berlin's opposition to intolerance and prejudice, especially fanatical monism, stereotypes and aggressive national­ism. It was to have appeared here for the first time, but it spoke so clearly to the events of i i September 2001 that I published it in the first issue of the New York Review of Books to appear thereafter.[12]

Berlin and his Critics

A_s Berlin indicates in 'Final Retrospect', the literature stimulated by the two central essays in Four Essays on Liberty has been large. Indeed, the rate of growth of the secondary literature has increased rather than diminishing as the years have passed. I have attempted to keep a tally of it on the official website of the Isaiah Berlin Literary Trust,¹ and I hope this resource will continue to be updated. The publication of Liberty provides an opportunity to supplement this bare list with a brief critical vade-mecum that will assist readers to find their way through the growing volume of articles and books discussing Berlin's ideas: the main focus, given this book's rationale, is on the discussion of liberty. This guide - beyond the capability of a mere editor - has kindly been provided by Dr Ian Harris of Leicester University, himself the author of a valuable article on 'Two Concepts'.²

Index

The index to Four Essays on Liberty is somewhat unsatisfactory - adequate in its coverage of names but too sparing of concepts. Accordingly, I have turned again to Douglas Matthews, sometime Librarian of the London Library, and faithful indexer of almost all my collections of Berlin's work, and invited him to start again from scratch.

Sources and acknowledgements

The original publication details of the pieces included in Liberty are as follows:

Five Essays on Liberty

Introduction: in Four Essays on Liberty (London and New York, 1969:

Oxford University Press) 'Political Ideas in the Twentieth Century': Foreign Affairs 28 No 3 (April 1950)

adjustments not included in this volume, where Berlin's manuscript is reproduced in a direct transcript (underlinings are indicated by italics), with only tiny corrections of slips of the pen. I have omitted material relevant only to the specific occasion in question.

¹ See p. ix above, note 2. ¹ See p. 357 below, note 14.

'Historical Inevitability': delivered on 12 May 1953 under the title 'History as an Alibi'' at the London School of Economics and Political Science as the first Auguste Comte Memorial Trust Lecture (London, 1954: Oxford University Press); repr. in Auguste Comte Memorial Lectures 1953-1962 (London, 1964: Athlone Press)

'Two Concepts of Liberty': Inaugural Lecture as Chichele Professor of Social and Political Theory, Oxford University, delivered on 3 I October 1958 (Oxford, 1958: Clarendon Press) 'John Stuart Mill and the Ends of Life': Robert Waley Cohen Memorial Lecture 1959, delivered on 2 December 1959 at County Hall, London (London, 1959: Council of Christians and Jews) 'From Hope and Fear Set Free': Presidential Address to the Aristote­lian Society, delivered on 14 October 1963 at 21 Bedford Square, London wci, Proceedings of the Aristotelian Society 64 (1963-4)

Other writings on liberty

'Liberty': in Ted Honderich (ed.), The Oxford Companion to Philoso­phy (Oxford, 1995: Oxford University Press); repr. in Berlin's The Power of Ideas (London, 200o:Chatto and Windus; Princeton, 2000: Princeton University Press) 'The Birth of Greek Individualism': as 'A Turning-Point in Political

Thought', Common Knowledge 7 No 3 (Winter 1998) 'Final Retrospect': excerpts from 'My Intellectual Path', published with 'The Purpose Justifies the Ways' as 'The First and the Last', New York Review of Books, 14 May 1988; repr. in The First and the Last (New York, 1999, New York Review of Books; London, 1999: Granta), and in The Power of Ideas (see under 'Liberty' above)

Autobiographical appendices

'The Purpose Justifies the Ways' (1922): published with 'My Intellec­tual Path' as 'The First and the Last' (see under 'Final Retrospect' above), without the editorial adjustments introduced for the present volume (see p. xxix above, note i) A Letter to George Kennan ( 195 I): published here for the first time 'Notes on Prejudice' (1981): see p. xxx above, note 3

I am grateful to Tim Barton, my successor at OUP, for allowing me to revisit the files on Four Essays, though he knew that I was not certain to be unfailingly diplomatic about what I found there. His colleagues Angela Griffin and Jo Stanbridge have shown great professionalism, courtesy and restraint as the book has assumed physical form. I should like to repeat my thanks to Roger

' At one stage, in proof, it was called 'History as the Culprit'.

Hausheer, Leofranc Holford-Strevens (who also contributed else­where) and Christopher Taylor for help in preparing the somewhat problematic text of 'The Birth of Greek Individualism' for publica­tion. Help with individual problems was kindly provided by Chimen Abramsky, Terrell Carver, Joshua Cherniss, Timothy Day, Steffen GroB, Roger Hausheer (who remains a perpetual, patient source of sage counsel), Jeremy Jennings, Leszek Kola- kowski, Mary Pickering, Hans Poser, Helen Rappaport, Mario Ricciardi, Philip Schofield, Marshall Shatz, Steven B. Smith and Manfred Steger. Betty Colquhoun keyboarded the whole book over a period of years with her usual exemplary dependability, and Serena Moore has masterminded the subsequent administrative processes, as well as suggesting several editorial improvements and an excellent metaphor. Samuel Guttenplan has supplied moral support and sensible advice, as well as providing, with Jennifer, a haven where the back of the editorial work was finally broken. Wolfson College and my generous benefactors continue to under­pin everything I do.

HENRY HARDY

Wolfson College, Oxford La Taillede, Laguepie 22 September 2oor

John Stallworthy, Esq., Oxford University Press, Ely House, 57* Dover Street, London, W.l.

8th February, 1968.

Dear Mr. Stallworthy,

I see that gradually but inexorably I am becomming if not your most intolerable (though I may be that too) certainly your most time consuming author. At the risk of inflicting a blow upon you which may seriously endanger your health such health and optimisim as you may have regained during your recent holiday I proposeto inflict yet ^^tb^. hideoQ8 blow upon you. Before I do this however let me assure you that I have carefully gone <^reough your most valuable list of queries and answered them all. I enclose the answers on separate sheets. I enclose a copy of your questionaire but the answers have proved somewhat more extensive than could comfortably be accommodated on it. Hence the extra sheets which I hope will not be a nusience.

The blow is this: it has been represented to me by kind friends (for once genuinely kind) that the book might be improved by the inclusion in it of yet another essay on the в^до subject namely my Presidential Address to the Ariscotelian Society a few years ago the title of which was "From Hope and Fear Set Free". This would make a fifth essay in the book and the title could be altered from "Four Essays on Liberty" either to "Five Essays on Liberty" or simply "Essays on Liberty", since five essays perhaps begin to deserve that title. The piece in question is not the worst that I have written, and I should like it included. To ease the birth pangs I am prepared to make the following promises:

Contd

The source of the title 'Five Essays on Liberty' (typed by one of Berlin's less accomplished secretaries)

FIVE ESSAYS ON LIBERTY

INTRODUCTION

. . . l'on immole a l'etre abstrait les etres reels; l'on offre au peuple en masse l'holocauste du peuple en detail.

Benjamin Constant, De I'esprit de conquete* I

The first of these five essays appeared in the mid-century number of the New York periodical Foreign Affairs; the remaining four originated in lectures. They deal with various aspects of individual liberty. They are concerned in the first place with the vicissitudes of this notion during the ideological struggles of our century; secondly, with the meaning it is given in the writings of historians, social scientists, and writers who examine the presuppo­sitions and methods of history or sociology; thirdly, with the importance of two major conceptions of liberty in the history of ideas; fourthly, with the part played by the ideal of individual liberty in the outlook of one of its most devoted champions, John Stuart Mill; and, finally, with the relationship between knowledge and freedom.

The first, fourth and fifth of these essays evoked little comment. The second and third stimulated wide and, as it seems to me, fruitful controversy. Since some of my opponents have advanced objections that seem to me both relevant and just, I propose to make it clear where I think that I stand convicted of mistakes or obscurities; other strictures (as I hope to show) seem to me mistaken. Some of my severest critics attack my views without adducing either facts or arguments, or else impute to me opinions that I do not hold; and even though this may at times be due to my own lack of clarity, I do not feel obliged to discuss, still less to defend, positions which, in some cases, appear to me as absurd as they do to those who assail them.[13]

The main issues between my serious critics and myself may be reduced to four heads: firstly, determinism and its relevance to our notions of men and their history; secondly, the place of value judgements, and, in particular, of moral judgements, in historical and social thinking; thirdly, the possibility and desirability of distinguishing, in the realm of political theory, what modern writers have called 'positive' liberty from 'negative' liberty, and the relevance of this distinction to the further difference between liberty and the conditions of liberty, as well as the question of what it is that makes liberty, of either sort, intrinsically worth pursuing or possessing; and finally, the issue of monism, of the unity or harmony of human ends. It seems to me that the unfavourable contrast sometimes drawn between 'negative' liberty and other, more obviously positive, social and political ends sought by men - such as unity, harmony, peace, rational self-direction, justice, self- government, order, co-operation in the pursuit of common pur­poses - has its roots, in some cases, in an ancient doctrine according to which all truly good things are linked to one another in a single, perfect whole; or, at the very least, cannot be incompatible with one another. This entails the corollary that the realisation of the pattern formed by them is the one true end of all rational activity, both public and private. If this belief should turn out to be false or incoherent, this might destroy or weaken the basis of much past and present thought and activity; and, at the very least, affect conceptions of, and the value placed on, personal and social liberty. This issue, too, is therefore both relevant and fundamental.

Let me begin with the most celebrated question of all as it affects human nature: that of determinism, whether causal or teleological. My thesis is not, as has been maintained by some of my most vehement critics, that it is certain (still less that I can show) that determinism is false; only that the arguments in favour of it are not conclusive; and that if it ever becomes a widely accepted belief and

' While I have not altered the text in any radical fashion, I have made a number of changes intended to clarify some of the central points which have been misunderstood by critics and reviewers. I am most grateful to Stuart Hampshire, H. L. A. Hart, Thomas Nagel and Patrick Gardiner for drawing my attention to errors and obscurities. I have done my best to remedy these, without, I feel sure, fully satisfying these distinguished and helpful critics.

enters the texture of general thought and conduct, the meaning and use of certain concepts and words central to human thought would become obsolete or else have to be drastically altered. This entails the corollary that the existing use of these basic words and concepts constitutes some evidence, not, indeed, for the proposi­tion that determinism is false, but for the hypothesis that many of those who profess this doctrine seldom, if ever, practise what they preach, and (if my thesis is valid) seem curiously unaware of what seems, prima facie, a lack of correspondence between their theory and their real convictions, as these are expressed in what they do and say. The fact that the problem of free will is at least as old as the Stoics; that it has tormented ordinary men as well as profes­sional philosophers; that it is exceptionally difficult to formulate clearly; that medieval and modern discussions of it, while they have achieved a finer analysis of the vast cluster of the concepts involved, have not in essentials brought us any nearer a definitive solution; that while some men seem naturally puzzled by it, others look upon such perplexity as mere confusion, to be cleared away by some single powerful philosophical solvent - all this gives determinism a peculiar status among philosophical questions.

I have, in these essays, made no systematic attempt to discuss the problem of free will as such; my focus is on its relevance to the idea of causality in history. Here I can only restate my original thesis that it seems to me patently inconsistent to assert, on the one hand, that all events are wholly determined to be what they are by other events (whatever the status of this proposition),¹ and, on the other, that men are free to choose between at least two possible courses of action - free not merely in the sense of being able to do what they choose to do (and because they choose to do it), but in the sense of not being determined to choose what they choose by causes outside their control. If it is held that every act of will or ^choice is fully determined by its respective antecedents, then (despite all that has been said against this) it still seems to me that this belief is incompatible with the notion of choice held by ordinary men, and by philosophers when they are not consciously defending a determinist position. More particularly, I see no way

' It has the appearance of a universal statement about the nature of things. But it can scarcely be straightforwardly empirical, for what item of experience would count as evidence against it?

round the fact that the habit of giving moral praise and blame, of congratulating and condemning men for their actions, with the implication that they are morally responsible for them, since they could have behaved differently, that is to say, need not have acted as they did (in some sense of 'could' and 'need' which is not purely logical or legal, but in which these terms are used in ordinary empirical discourse by both men in the street and historians), would be undermined by belief in determinism. No doubt the same words could still be used by determinists to express admira­tion or contempt for human characteristics or acts; or to encourage or deter; and such functions may be traceable to the early years of human society. However that may be, without the assumption of freedom of choice and responsibility in the sense in which Kant used these terms, one, at least, of the ways in which they are now normally used is, as it were, annihilated.

Determinism clearly takes the life out of a whole range of moral expressions. Very few defenders of determinism have addressed themselves to the question of what this range embraces and (whether or not this is desirable) what the effect of its elimination on our thought and language would be. Hence I believe that those historians or philosophers of history who maintain that responsi­bility and determinism are never incompatible with one another are mistaken, whether or not some form of determinism is true;[14] and again, whether or not some form of belief in the reality of moral responsibility is justified, what seems clear is that these possibilities are mutually exclusive: both beliefs may be groundless but both cannot be true. I have not attempted to adjudicate between these alternatives; only to maintain that men have, at all times, taken freedom of choice for granted in their ordinary discourse. And I further argue that if men became truly convinced that this belief was mistaken, the revision and transformation of the basic terms and ideas that this realisation would call for would be greater and more upsetting than the majority of contemporary determinists seem to realise. Beyond this I did not go, and do not propose to go now.

The belief that I undertook to demonstrate that determinism is false - on which much criticism of my argument has been based - is unfounded. I am obliged to say this with some emphasis, since some of my critics (notably E. H. Carr) persist in attributing to me a claim to have refuted determinism. But this, like another odd view ascribed to me, namely that historians have a positive duty to moralise, is a position that I have never defended or held; this is a point to which I shall have occasion to revert later. More specifi­cally, I have been charged with confusing determinism with fatalism.[15] But this, too, is a complete misunderstanding. I assume that what is meant or implied by fatalism is the view that human decisions are mere by-products, epiphenomena, incapable of influ­encing events which take their inscrutable course independently of human wishes. I have never attributed this unplausible position to any of my opponents. The majority of them cling to 'self- determinism' - the doctrine according to which men's characters and 'personality structures' and the emotions, attitudes, choices, decisions and acts that flow from them do indeed play a full part in what occurs, but are themselves results of causes, psychical and physical, social and individual, which in turn are effects of other causes, and so on, in unbreakable sequence. According to the best- known version of this doctrine, I am free if I can do what I wish and, perhaps, choose which of two courses of action I shall take. But my choice is itself causally determined; for if it were not, it would be a random event; and these alternatives exhaust the possibilities; so that to describe choice as free in some further sense, as neither caused nor random, is to attempt to say something meaningless. This classical view, which to most philosophers appears to dispose of the problem of free will, seems to me simply a variant of the general determinist thesis, and to rule out responsibility no less than its 'stronger' variant. Such 'self- determinism' or 'weak determinism', in which, since its original formulation by the Stoic sage Chrysippus, many thinkers have come to rest, was described by Kant as a 'miserable subterfuge'.[16]

William James labelled it 'soft determinism', and called it, perhaps too harshly, 'a quagmire of evasion'.¹

I cannot see how one can say of Helen not only that hers was the face that launched a thousand ships but, in addition, that she was responsible for (and did not merely cause) the Trojan War, if the war was due solely to something that was the result not of a free choice - to elope with Paris - which Helen need not have made, but only of her irresistible beauty. Sen, in his clear and moderately worded criticism, concedes what some of his allies do not - that there is an inconsistency between, at any rate, some meanings attached to the contents of ordinary moral judgement on the one hand, and determinism on the other. He denies, however, that belief in determinism need eliminate the possibility of rational moral judgement, on the ground that such judgements could still be used to influence men's conduct, by acting as stimuli or deterrents. In somewhat similar terms, Ernest Nagel, in the course of a characteristically scrupulous and lucid argument/ says that, even on the assumption of determinism, praise, blame and assump­tion of responsibility generally could affect human behaviour - for example, by having an effect on discipline, effort and the like, whereas they would (presumably) not in this way affect a man's digestive processes or the circulation of his blood.

This may be true but it does not affect the central issue. Our value judgements - eulogies or condemnations of the acts or characters of men dead and gone - are not intended solely, or even primarily, to act as utilitarian devices, to encourage or warn our contemporaries, or as beacons to posterity. When we speak in this way we are not attempting merely to influence future action (though we may, in fact, be doing this too) or solely to formulate quasi-aesthetic judgements - as when we testify to the beauty or ugliness, intelligence of stupidity, generosity or meanness of others (or ourselves) - attributes which we are then simply attempting to grade according to some scale of values. If someone praises or condemns me for choosing as I did, I do not always say either 'This is how I am made; I cannot help behaving so', or 'Please go on

' William James, 'The Dilemma of Determinism': p. 149 in his The Will to Believe, and Other Essays in Popular Philosophy (New York etc., 1897).

^г The Structure of Science: Problems in the Logic of Scientific Explanation (London, 1961), pp. 599-605. Also, by the same author, 'Determinism in History', Philosophy and Phenomenological Research 20 (19 5 9-60), 29 i -3 i 7, at 3 I 1-16.

saying this, it is having an excellent effect on me: it strengthens [or weakens] my resolution to go to war, or to join the Communist Party.'

It may be that such words, like the prospect of rewards and punishments, do affect conduct in important ways, and that this makes them useful or dangerous. But this is not the point at issue. It is whether such praise, blame and so on are merited, morally appropriate, or not. One can easily imagine a case where we think that a man deserves blame, but consider that to utter it may have a bad effect, and therefore say nothing. But this does not alter the man's desert, which, whatever its analysis, entails that the agent could have chosen, and not merely acted, otherwise. If I judge that a man's conduct was in fact determined, that he could not have behaved (felt, thought, desired, chosen) otherwise, I should regard this kind of praise or blame as inappropriate to his case. If determinism is true, the concept of merit or desert, as these are usually understood, has no application. If all things and events and persons are determined, then praise and blame do indeed become purely pedagogical devices - hortatory and minatory. Or else they are quasi-descriptive - they grade in terms of distance from some ideal. They comment on the quality of men, what men are and can be and do, and may themselves alter it and, indeed, be used as deliberate means towards it, as when we reward or punish an animal; save that in the case of men we assume the possibility of communication with them, which we cannot do in the case of animals.

This is the heart of 'soft' determinism - the so-called Hobbes- Hume-Schlick doctrine. If, however, the notions of desert, merit, responsibility and so forth rested on the notion of choices not themselves fully caused, they would, on this view, turn out to be irrational or incoherent; and would be abandoned by rational men. The majority of Spinoza's interpreters suppose him to have ^maintained precisely this, and a good many of them think that he was right. But whether or not Spinoza did in fact hold this view, and whether or not he was right in this respect, my thesis is that, however it may be with Spinoza, most men and most philosophers and historians do not speak or act as if they believe this. For if the determinist thesis is genuinely accepted, it should, at any rate to men who desire to be rational and consistent, make a radical difference. Sen, with admirable consistency and candour, does indeed explain that, when determinists use the language of moral praise and blame, they are like atheists who still mention God, or lovers who speak of being faithful 'till the end of time';¹ such talk is hyperbolic and not meant to be taken literally. This does at least concede (as most determinists do not) that if these words were taken literally something would be amiss. For my part I see no reason for supposing that most of those who use such language, with its implication of free choice among alternatives, whether in the future or in the past, mean this not literally, but in some Pickwickian or metaphorical or rhetorical way. Ernest Nagel points out that determinists, who, like Bossuet, believed in the omnipotence and omniscience of Providence and its control over every human step, nevertheless freely attributed moral responsibil­ity to individuals; and that adherents of determinist faiths - Muslims, Calvinists and others - have not refrained from attribu­tion of responsibility and a generous use of praise and blame.² Like much that Ernest Nagel says, this is perfectly true.³ But it is nothing to the issue: the fact that not all human beliefs are coherent is not novel. These examples merely point to the fact that men evidently find it perfectly possible to subscribe to determinism in the study and disregard it in their lives. Fatalism has not bred passivity in Muslims, nor has determinism sapped the vigour of Calvinists or Marxists, although some Marxists feared that it might. Practice sometimes belies profession, no matter how sincerely held.

E. H. Carr goes a good deal further. He declares: 'The fact is that all human actions are both free and determined, according to the

' op. cit. (p. 7 above, note i), p. i 14.

² The Structure of Science (see p. 8 above, note 2), pp. 603-4.

¹ See also a similar but equally unconvincing argument in the inaugural lecture by Sydney Pollard at Sheffield University, 'Economic History - A Science of Society?', Past and Present 30 (April 1965), 3-22. Much of what Pollard says seems to me valid and worth saying, but his view, supported by an appeal to history, that men's professions must be consistent with their practice is to say the least oddly surprising in a historian. Nagel (The Structure of Science, p. 602) suggests that belief in free will may relate to determinism much as the conviction that a table has a hard surface relates to the hypothesis that it consists of whirling electrons; the two descriptions answer questions at different levels, and therefore do not clash. This does not seem to me an apt parallel. To believe that the table is hard, solid, at rest etc. entails no beliefs about electrons; and is, in principle, compatible with any doctrine about them: the levels do not touch. But if I supposed a man to have acted freely, and am later told that he acted as he did because he was 'made that way', and could not have acted differently, I certainly suppose that something that I believed is being denied.

point of view from which one considers them.' And again: 'adult human beings are morally responsible for their own personality'/ This seems to me to present the reader with an insoluble puzzle. If Carr means that human beings can transform the nature of their personality, while all antecedents remain the same, then he denies causality; if they cannot, and acts can be fully accounted for by character, then talk of responsibility (in the ordinary sense of this word, in which it implies moral blame) does not make sense. There _are no doubt many senses of 'can'; and much light has been shed on this by important distinctions made by acute modern philoso­phers. Nevertheless, if I literally cannot make my character or behaviour other than it is by an act of choice (or a whole pattern of such acts) which is itself not fully determined by causal antece­dents, then I do not see in what normal sense a rational person could hold me morally responsible either for my character or for my conduct. Indeed the notion of a morally responsible being becomes, at best, mythological; this fabulous creature joins the ranks of nymphs and centaurs.

The horns of this dilemma have been with us for over two millennia, and it is useless to try to escape or soften them by the comfortable assertion that it all depends on the point of view from which we regard the question. This problem, which preoccupied Mill (and to which, in the end, he returned so confused an answer), and from the torment of which William James escaped only as a result of reading Renouvier, and which is still well to the forefront of philosophical attention, cannot be brushed aside by saying that the questions to which scientific determinism is the answer are different from those which are answered by the doctrine of voluntarism and freedom of choice between alternatives; or that the two types of question arise at different 'levels', so that a pseudo- problem has arisen from the confusion of these 'levels' (or the corresponding categories). The question to which determinism and indeterminism, whatever their obscurities, are the rival answers is one and not two. What kind of question it is - empirical, conceptual, metaphysical, pragmatic, linguistic - and what schema or model of man and nature is implicit in the terms used are major

' Edward Hallett Carr, What is History? (London, 1961), p. 89 (p. 95 in the paperback edition, Harmondsworth, 1964; page references to this edition are added in parentheses in subsequent notes).

philosophical issues; but it would be out of place to discuss them here.

Nevertheless, if only because some of the sharpest criticisms of my thesis come from philosophers concerned with this central issue, it cannot be entirely passed over. Thus J. A. Passmore[17]urges two considerations against me: (a) That the concept of Laplace's observer, who can infallibly predict the future, since he has all the relevant knowledge of antecedent conditions and laws that he needs for this, cannot in principle be formulated, because the notion of an exhaustive list of all the antecedents of an event is not coherent; we can never say of any state of affairs 'These are all the antecedents there are; the inventory is complete.' This is clearly true. Nevertheless, even if determinism were offered as no more than a pragmatic policy - 'I intend to act and think on the assumption that every event has an identifiable sufficient cause or causes' - this would satisfy the determinist's demand. Yet such a resolve would make a radical difference, for it would effectively take the life out of any morality that works with such notions as responsibility, moral worth and freedom in Kant's sense, and do so in ways and with logical consequences which determinists as a rule either forbear to examine, or else play down.

(b) That the more we find out about a prima facie morally culpable act, the more we are likely to realise that the agent, given the particular circumstances, characters, antecedent causes involved, was prevented from taking the various courses which we think he should have adopted; we condemn him too easily for failing to do or be what he could not have done or been. Ignorance, insensitiveness, haste, lack of imagination darken counsel and blind us to the true facts; our judgements are often shallow, dogmatic, complacent, irresponsible, unjust, barbarous. I sympathise with the humane and civilised considerations which inspire Passmore's verdict. Much injustice and cruelty has sprung from avoidable ignorance, prejudice, dogma and lack of understanding. Neverthe­less to generalise this - as Passmore seems to me to do - is to fall into the old tout comprendre fallacy in disarmingly modern dress. If (as happens to those who are capable of genuine self-criticism) the more we discover about ourselves the less we are inclined to forgive ourselves, why should we assume that the opposite is valid for others, that we alone are free, while others are determined? To expose the deleterious consequences of ignorance or irrationality is one thing; to assume that these are the sole sources of moral indignation is an illicit extrapolation; it would follow from Spinoz- ist premisses, but not necessarily from others. Because our judge­ments about others are often superficial or unfair, it does not follow that one must never judge at all; or, indeed, that one can avoid doing so. As well forbid all men to count, because some cannot add correctly.

Morton White attacks my contentions from a somewhat differ­ent angle.¹ He concedes that one may not, as a rule, condemn (as being 'wrong') acts which the agent could not help perpetrating (for example, Booth's killing of Lincoln, on the assumption that he was caused to choose to do this, or anyway to do it whether or not he so chose). Or at least White thinks that it is unkind to blame a man for a causally determined action; unkind, unfair, but not inconsistent with determinist beliefs. We could, he supposes, conceive of a culture in which such moral verdicts would be normal. Hence it may be mere parochialism on our part to assume that the discomfort we may feel in calling causally determined acts right or wrong is universal, and springs from some basic category which governs the experience of all possible societies.

White discusses what is implied by calling an act 'wrong'. I am concerned with such expressions as 'blameworthy', 'something you should not have done', 'deserving to be condemned' - none of which is equivalent to 'wrong' or, necessarily, to each other. But even so, I wonder whether White, if he met a kleptomaniac, would think it reasonable to say to him: 'You cannot, it is true, help choosing to steal, even though you may think it wrong to do so. Nevertheless you must not do it. Indeed, you should choose to refrain from it. If you go on, we shall judge you not only to be a wrongdoer, but to deserve moral blame. Whether this deters you or not, you will deserve it equally in either case.' Would White not feel that something was seriously amiss about such an approach, and that not merely in our own society, but in any world in which such moral terminology made sense? Or would he think this very question to be evidence of insufficient moral imagination in the

' Foundations of Historical Knowledge (New York and London, 1965), pp. 275 ff. I cannot pretend to be able to do justice here to the complex and interesting thesis which White's luminous book expounds. I ask him to forgive me for the summary character of this brief rejoinder.

questioner? Is it merely unkind or unfair to reproach men with what they cannot avoid doing, or, like much cruelty and injustice, irrational too? If you said to a man who betrayed his friends under torture that he should not have done this, that his act was morally wrong, even though you are convinced that, being what he is, he could not help choosing to act as he did, could you, if pressed, give reasons for your verdict? What could they be? That you wished to alter his (or others') behaviour in the future? Or that you wished to ventilate your revulsion?

If this were all, questions of doing him justice would not enter at all. Yet if you were told that in blaming a man you were being unfair or wickedly blind, because you had not troubled to examine the difficulties under which the man laboured, the pressures upon him, and so on, this kind of reproach rests on the assumption that in some cases, if not in all, the man could have avoided the choice that you condemn, only a good deal less easily than you realise, at the price of martyrdom, or the sacrifice of the innocent, or at some cost which your critic believes that you, the moraliser, have no right to demand. Hence the critic rightly reproaches you for culpable ignorance or inhumanity. But if you really thought that it was (causally) impossible for the man to have chosen what you would have preferred him to choose, is it reasonable to say that he should nevertheless have chosen it? What reasons can you, in principle, adduce for attributing responsibility or applying moral rules to him (such as Kant's maxims, which we understand whether or not we accept them) which you would not think it reasonable to apply in the case of compulsive choosers - kleptomaniacs, dipso­maniacs and the like? Where would you draw the line, and why?

If the choices in all these cases are causally determined, however different the causes, in some cases being compatible (or, according to some views, identical) with the use of reason, in others not, why is it rational to blame in one case and not in the other? I exclude the utilitarian argument for praise or blame or threats or other incentives, since White, rightly in my view, ignores it too, to concentrate on the moral quality of blame. I cannot see why it is less unreasonable (and not merely less futile) to blame a man psychologically unable to refrain from it for acting cruelly than a physical cripple for possessing a deformed limb. To condemn a murderer is no more or less rational than to blame his dagger; so reasoned Godwin. At least he was consistent in his fanatical way. Although his best-known book is called Political Justice, it is not easy to tell what justice, as a moral concept, would mean to a convinced determinist. I could grade just and unjust acts, like legal and illegal ones, like ripe and unripe peaches. But if a man could not help acting as he did, how much would it mean to say that something 'served him right'? The notion of poetic justice, of just deserts, of moral desert as such, would, if this were the case, not merely have no application, but become scarcely intelligible.

When Samuel Butler in Erewhon makes crimes objects of sympathy and pity, but ill health an offence which leads to sanctions, he is set on emphasising not the relativity of moral values, but their irrationality in his own society - the irrationality of blame directed at moral or mental aberrations, but not to physical or physiological ones. I know of no more vivid way of bringing out how different our moral terminology and conduct would be if we were the really consistent scientific determinists that some suppose that we ought to be. The more rigorous sociological determinists do indeed say precisely this, and consider that not only retribution or revenge, but justice too - outside its strictly legal sense - conceived as a moral standard or canon not determined by alterable rules, is a pre-scientific notion grounded in psychological immaturity and error. As against this Spinoza and Sen seem to me to be right. There are some terms which, if we took determinism seriously, we should no longer use, or use only in some peculiar sense, as we speak of witches or the Olympian Gods. Such notions as justice, equity, desert, fairness would certainly have to be re-examined if they were to be kept alive at all and not relegated to the role of discarded figments - fancies rendered harmless by the march of reason, myths potent in our irrational youth, exploded, or at any rate rendered innocuous, by the progress of scientific knowledge. If determinism is valid, this is a price that we must pay. Whether or not we are ready to do so, let this prospect at least be faced.

If our moral concepts belong only to our own culture and society, then what we should be called upon to say to a member of White's unfamiliar culture is not that he was logically contradicting himself in professing determinism and yet continuing to utter or imply moral judgements of a Kantian sort, but that he was being incoherent, that we could not see what reasons he could have for using such terms, that his language, if it was intended to apply to the real world, was no longer sufficiently intelligible to us. Of course the fact that there have been, and no doubt may still be, plenty of thinkers, even in our own culture, who at one and the same time profess belief in determinism, and yet do not feel in the least inhibited from dispensing this kind of moral praise and blame freely, and pointing out to others how they should have chosen, shows only, if I am right, that some normally lucid and self-critical thinkers are at times liable to confusion. My case, in other words, amounts to making explicit what most men do not doubt - namely that it is not rational both to believe that choices are caused, and to consider men as deserving of reproach or indignation (or their opposites) for choosing to act or refrain as they do.

The supposition that, if determinism were shown to be valid, ethical language would have to be drastically revised is not a psychological or a physiological, still less an ethical, hypothesis. It is an assertion about what any system of thought that employs the basic concepts of our normal morality would permit or exclude. The proposition that it is unreasonable to condemn men whose choices are not free rests not on a particular set of moral values (which another culture might reject) but on the particular nexus between descriptive and evaluative concepts which governs the language we use and the thoughts we think. To say that you might as well morally blame a table as an ignorant barbarian or an incurable addict is not an ethical proposition, but one which emphasises the conceptual truth that this kind of praise and blame makes sense only among persons capable of free choice. This is what Kant appeals to; it is this fact that puzzled the early Stoics; before them freedom of choice seems to have been taken for granted; it is presupposed equally in Aristotle's discussion of voluntary and involuntary acts and in the thinking of unphilosoph- ical persons to this day.

One of the motives for clinging to determinism seems to be the fear on the part of the friends of reason that it is presupposed by scientific method as such. Thus Stuart Hampshire tells us that:

In the study of human behaviour, philosophical superstition might now easily take over the role of traditional religious superstitions as an obstruction to progress. In this context superstition is a confusion of the belief that human beings ought not to be treated as if they were natural objects with the belief that they are not in reality natural objects: one may so easily move from the moral proposition that persons ought not to be manipulated and controlled, like any other natural objects, to the different, and quasi-philosophical, proposition that they cannot be manipulated and controlled like any other natural objects. In the present climate of opinion a very natural fear of planning and social technology is apt to be dignified as a philosophy of indeterminism.'

This strongly worded cautionary statement seems to me charac­teristic of the widespread and influential feeling I have mentioned that science and rationality are in danger if determinism is rejected or even doubted. This fear appears to me to be groundless; to do one's best to find quantitative correlations and explanations is not to assume that everything is quantifiable; to proclaim that science is the search for causes (whether this is true or false) is not to say that all events have them. Indeed the passage that I have quoted seems to me to contain at least three puzzling elements.

We are told that to confuse 'the belief that human beings ought not to be treated as if they were natural objects with the belief that they are not in reality natural objects' is superstitious. But what other reason have I for not treating human beings 'as if they were natural objects' than my belief that they differ from natural objects in some particular respects - those in virtue of which they are human - and that this fact is the basis of my moral conviction that I should not treat them as objects, that is, solely as means to my ends, and that it is in virtue of this difference that I consider it wrong freely to manipulate, coerce, brainwash them and so on? If I am told not to treat something as a chair, the reason for this may be the fact that the object in question possesses some attribute which ordinary chairs do not, or has some special association for me or others which distinguishes it from ordinary chairs, a characteristic which might be overlooked or denied. Unless men are held to possess some attribute over and above those which they have in common with other natural objects - animals, plants, things - (whether this difference is itself called natural or not), the moral command not to treat men as animals or things has no rational foundation. I conclude that, so far from being a confusion of two different kinds of proposition, this connection between them cannot be severed without making at least one of them groundless; and this is certainly unlikely to forward the progress of which the author speaks.

As for the warning not to move from the proposition that 'persons ought not to be manipulated and controlled' to the

' 'Philosophy and Madness', Listener 78 (July-December 1967), 289-92, at 291.

proposition that 'they cannot be manipulated or controlled like any other natural object', it is surely more reasonable to suppose that if I tell you not to do it, that is not because I think persons cannot be so treated, but because I believe that it is all too likely that they can. If I order you not to control and manipulate human beings, it is not because I think that, since you cannot succeed, this will be a sad waste of your time and effort; but on the contrary, because I fear that you may succeed all too well, and that this will deprive men of their freedom, a freedom which, if they can only escape from too much control and manipulation, I believe they may be able to preserve.

(c) 'Fear of planning and social technology' may well be most acutely felt by those who believe that these forces are not irresistible; and that if men are not too much interfered with they will have opportunities of choosing freely between possible courses of action, not merely (as determinists believe) of, at best, imple­menting choices themselves determined and predictable. The latter may in fact be our actual condition. But if one prefers the former state - however difficult it may be to formulate - is this a superstition, or some other case of 'false consciousness'? It is such only if determinism is true. But this is a viciously circular argument. Could it not be maintained that determinism itself is a superstition generated by a false belief that science will be compro­mised unless it is accepted, and is therefore itself a case of 'false consciousness' generated by a mistake about the nature of science? Any doctrine could be turned into a superstition, but I do not myself see any reason for holding that either determinism or indeterminism is, or need turn into, one.[18]

To return to non-philosophical writers. The writings of those who have stressed the inadequacy of the categories of the natural sciences when applied to human action have so far transformed the question as to discredit the crude solutions of the nineteenth- and twentieth-century materialists and positivists. Hence all serious discussion of the issues must now begin by taking some account of the world-wide discussion of the subject during the last twenty- five years. When E. H. Carr maintains that to attribute historical events to the acts of individuals ('biographical bias') is childish, or at any rate childlike, and that the more impersonal we make our historical writing, the more scientific, and therefore mature and valid, it will be, he shows himself a faithful - too faithful - follower of the eighteenth-century dogmatic materialists. This doctrine no longer seemed altogether plausible even in the day of Comte and his followers, or, for that matter, of the father of Russian Marxism, Plekhanov, who, for all his brilliance, in his philosophy of history owed more to eighteenth-century materialism and nineteenth- century positivism than to Hegel or the Hegelian elements in Marx.

Let me give Carr his due. When he maintains that animism or anthropomorphism - the attribution of human properties to inanimate entities - is a symptom of a primitive mentality, I have no wish to controvert this. But to compound one fallacy with another seldom advances the cause of truth. Anthropomorphism is the fallacy of applying human categories to the non-human world. But then there presumably exists a region where human categories do apply: namely the world of human beings. To suppose that only what works in the description and prediction of non-human nature must necessarily apply to human beings too and that the categories in terms of which we distinguish the human from the non-human must therefore be delusive - to be explained away as aberrations of our early years - is the opposite error, the animist or anthropomor­phic fallacy stood on its head. What scientific method can achieve, it must, of course be used to achieve. Anything that statistical methods or computers or any other instrument or method fruitful in the natural sciences can do to classify, analyse, predict or 'retrodict' human behaviour should, of course, be welcomed; to refrain from using these methods for some doctrinaire reason would be mere obscurantism. However, it is a far cry from this to the dogmatic assurance that the more the subject-matter of an enquiry can be assimilated to that of a natural science the nearer the truth we shall come. This doctrine, in Carr's version, amounts to saying that the more impersonal and general, the more valid, the more generic, the more grown up; the more attention to individu­als, their idiosyncrasies and their role in history, the more fanciful, the remoter from objective truth and reality. This seems to me no more and no less dogmatic than the opposite fallacy - that history is reducible to the biographies of great men and their deeds. To assert that the truth lies somewhere between these extremes, between the equally fanatical positions of Comte and Carlyle, is a dull thing to say, but may nevertheless be closer to the truth. As an eminent philosopher of our time¹ once drily observed, there is no a priori reason for supposing that the truth, when it is discovered, will prove interesting. Certainly it need not prove startling or upsetting; it may or may not; we cannot tell. This is not the place to examine Carr's historiographical views, which seem to me to breathe the last enchantments of the Age of Reason, more rational­ist than rational, with all the enviable simplicity, lucidity and freedom from doubt or self-questioning which characterised this field of thought in its unclouded beginnings, when Voltaire and Helvetius were on their thrones; before the Germans, with their passion for excavating everything, ruined the smooth lawns and symmetrical gardens. Carr is a vigorous and enjoyable writer, touched by historical materialism, but essentially a late positivist, in the tradition of Auguste Comte, Herbert Spencer and H. G. Wells; what Sainte-Beuve called 'un grand simplificateur'/ un­troubled by the problems and difficulties which have bedevilled the subject since Herder and Hegel, Marx and Max Weber. He is respectful towards Marx, but remote from his complex vision; a master of short ways and final answers to the great unanswered questions.

But if I cannot here attempt to deal with Carr's position with the care that it deserves, I can at least try to reply to some of his severest strictures on my own opinions. His gravest charges against me are threefold: (a) that I believe determinism to be false and reject the axiom that everything has a cause, which, according to Carr, 'is a condition of our capacity to understand what is going on around us'/ (b) that I 'insist with great vehemence that it is the duty of the historian "to judge Charlemagne or Napoleon or

' Identified by Berlin elsewhere (the wording varies) as C. I. Lewis. I have not been able to find this remark in Lewis's published writings. Ed.

The phrase 'grand simplificateur' and the word 'simplificateur' itself were coined by Sainte-Beuve to describe Benjamin Franklin in 'Franklin a Passy' (29 November 1 8 52): p. I 8 I in C.-A. Sainte-Beuve, Causenes du lundi (Paris, [1926-42]), vol. 7. (The equally familiar 'terribles simplificateurs', used by Berlin on p. 56 below, note 1 - and cf. p. 290 below - was coined by Jacob Burckhardt in a letter of 24 July 1889 to Friedrich von Preen.) Ed.

op. cit. (p. I 1 above, note 1), pp. 87-8 (93-4).

Genghis Khan or Hitler or Stalin for their massacres'", [19] that is, to award marks for moral conduct to historically important individu­als; (c) that I believe that explanation in history is an account in terms of human intentions, to which Carr opposes the alternative concept of 'social forces'.²

To all this I can only say once again: (a) I have never denied (nor considered) the logical possibility that some version of determin­ism may, in principle (although, perhaps, only in principle), be a valid theory of human conduct; still less do I hold myself to have refuted it. My sole contention has been that belief in it is not compatible with beliefs deeply embedded in the normal speech and thought of either ordinary men or historians, at any rate in the Western world; and therefore that to take determinism seriously would entail a drastic revision of these central notions - an upheaval of which neither Carr's nor any other historian's practice has, as yet, provided any conspicuous examples. I know of no conclusive argument in favour of determinism. But that is not my point; it is that the actual practice of its supporters, and their reluctance to face what unity of theory and practice in this case would cost them, indicate that such theoretical support is not at present to be taken too seriously, whoever may claim to provide it.

(b) I am accused of inviting historians to moralise. I do nothing of the kind. I merely maintain that historians, like other men, use language which is inevitably shot through with words of evaluative force, and that to invite them to purge their language of it is to ask them to perform an abnormally difficult and self-stultifying task. To be objective, unbiased, dispassionate is no doubt a virtue in historians, as in anyone who wishes to establish truth in any field. But historians are men, and are not obliged to attempt to dehumanise themselves to a greater degree than other men; the topics they choose for discussion, their distribution of attention and emphasis, are guided by their own scale of values, which, if they are either to understand human conduct or to communicate their vision to their readers, must not diverge too sharply from the common values of men.

To understand the motives and outlook of others it is not, of course, necessary to share them; insight does not entail approval; the most gifted historians (and novelists) are the least partisan;

some distance from the subject is required. But while comprehen­sion of motives, moral or social codes, entire civilisations does not require acceptance of, or even sympathy with, them, it does presuppose a view of what matters to individuals or groups, even if such values are found repulsive. And this rests on a conception of human nature, human ends, which enters into the historian's own ethical or religious or aesthetic outlook. These values, particularly the moral values which govern the selection of facts by historians, the light in which they exhibit them, are conveyed, and cannot but be conveyed, by their language as much and as little as are those of anyone else who seeks to understand and describe men. The criteria which we use in judging the work of historians are not, and need not, in principle, be, different from those by which we judge specialists in other fields of learning and imagination. In criticising the achievements of those who deal with human affairs we cannot sharply divorce 'facts' from their significance. 'Values enter into the facts and are an essential part of them. Our values are an essential part of our equipment as human beings.' These words are not mine. They are the words of none other (the reader will surely be astonished to learn) than Carr himself.[20] I might have chosen to formulate this proposition differently. But Carr's words are quite sufficient for me; on them I am content to rest my case against his charges.

There is clearly no need for historians formally to pronounce moral judgements, as Carr mistakenly thinks that I wish them to do. They are under no obligation as historians to inform their readers that Hitler did harm to mankind, whereas Pasteur did good (or whatever they may think to be the case). The very use of normal language cannot avoid conveying what the author regards as commonplace or monstrous, decisive or trivial, exhilarating or depressing. In describing what occurred I can say that so many million men were brutally done to death; or alternatively that they perished; laid down their lives; were massacred; or simply that the population of Europe was reduced, or that its average age was lowered; or that many men lost their lives. None of these descriptions of what took place is wholly neutral: all carry moral implications. What the historian says will, however careful he may be to use purely descriptive language, sooner or later convey his attitude. Detachment is itself a moral position. The use of neutral

' ibid., p. I 2 5 (13 i).

language ('Himmler caused many persons to be asphyxiated') conveys its own ethical tone.

I do not mean to say that severely neutral language about human beings is unattainable. Statisticians, compilers of intelligence reports, research departments, sociologists and economists of certain kinds, official reporters, compilers whose task it is to provide data for historians or politicians, can and are expected to approach it. But this is so because these activities are not autono­mous but are designed to provide the raw material for those whose work is intended to be an end in itself - historians, men of action. The research assistant is not called upon to select and emphasise what counts for much, and play down what counts for little, in human life. The historian cannot avoid this; otherwise what he writes, detached as it will be from what he, or his society, or some other culture, regards as central or peripheral, will not be history. If history is what historians do, then the central issue, which no historian can evade, whether he knows this or not, is how we (and other societies) come to be as we are or were. This, eo facto, entails a particular vision of society, of men's nature, of the springs of human action, of men's values and scales of value - something that physicists, physiologists, physical anthropologists, grammarians, econometricians or certain sorts of psychologists (like the pro­viders of data for others to interpret) may be able to avoid. History is not an ancillary activity; it seeks to provide as complete an account as it can of what men do and suffer; to call them men is to ascribe to them values that we must be able to recognise as such, otherwise they are not men for us. Historians cannot therefore (whether they moralise or not) escape from having to adopt some position about what matters and how much (even if they do not ask why it matters). This alone is enough to render the notion of a 'value-free' history, of the historian as a transcriber ipsis rebus dictantibus,' an illusion.

Perhaps this is all that Acton urged against Creighton: not simply that Creighton used artificially non-moral terms, but that in using them to describe the Borgias and their acts he was, in effect, going some way towards exonerating them; that, whether he was right or wrong to do it, he was doing it; that neutrality is also a moral attitude, and that it is as well to recognise it for what it is.

Acton had no doubt that Creighton was wrong. We may agree with Acton or with Creighton. But in either case we are judging, conveying, even when we prefer not to state, a moral attitude. To invite historians to describe men's lives but not the significance of their lives in terms of what Mill called the permanent interests of man, however conceived, is not to describe their lives. To demand of historians that they try to enter imaginatively into the experi­ence of others and forbid them to display moral understanding is to invite them to tell too small a part of what they know, and to deprive their work of human significance. This is in effect all that I have to say against Carr's moral sermon against the bad habit of delivering moral sermons.

No doubt the view that there exist objective moral or social values, eternal and universal, untouched by historical change, and accessible to the mind of any rational man if only he chooses to direct his gaze at them, is open to every sort of question. Yet the possibility of understanding men in one's own or any other time, indeed of communication between human beings, depends upon the existence of some common values, and not on a common 'factual' world alone. The latter is a necessary but not a sufficient condition of human intercourse. Those who are out of touch with the external world are described as abnormal or, in extreme cases, insane. But so also - and this is the point - are those who wander too far from the common public world of values. A man who declares that he once knew the difference between right and wrong, but has forgotten it, will scarcely be believed; if he is believed, he is rightly considered deranged. But so too is a man who does not merely approve or enjoy or condone, but literally cannot grasp what conceivable objection anyone could have to, let us suppose, a rule permitting the killing of any man with blue eyes, with no reason given. He would be considered about as normal a specimen of the human race as one who cannot count beyond six, or thinks it probable that he is Julius Caesar. What such normative (not descriptive) tests for insanity rest on is what gives such plausibility as they have to doctrines of natural law, particularly in versions which refuse them any a priori status. Acceptance of common values (at any rate some irreducible minimum of them) enters our conception of a normal human being. This serves to distinguish such notions as the foundations of human morality on the one hand from such other notions as custom, or tradition, or law, or manners, or fashion, or etiquette - all those regions in which the occurrence of wide social and historical, national and local differ­ences and change are not regarded as rare or abnormal, or evidence of extreme eccentricity or insanity, or indeed as undesirable at all; least of all as philosophically problematic.

No historical writing which rises above a bare chronicler's narrative, and involves selection and the unequal distribution of emphasis, can be wholly wertfrei} What then distinguishes moralising that is justly condemned from that which seems unescapable from any degree of reflection on human affairs? Not its overtness: mere choice of apparently neutral language can well seem, to those who do not sympathise with an author's views, even more insidious. I have attempted to deal with what is meant by bias and partiality in the essay on historical inevitability. I can only repeat that we seem to distinguish subjective from objective appraisal by the degree to which the central values conveyed are those which are common to human beings as such, that is, for practical purposes, to the great majority of men in most places and times. This is clearly not an absolute or rigid criterion; there is variation, there are virtually unnoticeable (as well as glaring) national, local and historical peculiarities, prejudices, superstitions, rationalisations and their irrational influence. But neither is this criterion wholly relative or subjective, otherwise the concept of man would become too indeterminate, and men or societies, divided by unbridgeable normative differences, would be wholly unable to communicate across great distances in space and time and culture.

Objectivity of moral judgement seems to depend on (almost consist in) the degree of constancy in human responses. This notion cannot in principle be made sharp and unalterable. Its edges remain blurred. Moral categories - and categories of value in general - are nothing like as firm and ineradicable as those of, say, the perception of the material world, but neither are they as relative or as fluid as some writers have too easily, in their reaction against the dogmatism of the classical objectivists, tended to assume. A minimum of common moral ground - interrelated concepts and categories - is intrinsic to human communication. What they are, how flexible, how far liable to change under the impact of what 'forces' - these are empirical questions, in a region claimed by

' 'Value-free.' moral psychology and historical and social anthropology, fascinat­ing, important and insufficiently explored. To demand more than this seems to me to wish to move beyond the frontiers of communicable human knowledge.

(c) I am accused of supposing that history deals with human motives and intentions, for which Carr wishes to substitute the action of 'social forces'. To this charge I plead guilty. I am obliged to say once more that anyone concerned with human beings is committed to consideration of motives, purposes, choices, the specifically human experience that belongs to human beings uniquely, and not merely with what happened to them as animate or sentient bodies. To ignore the play of non-human factors; or the effect of the unintended consequences of human acts; or the fact that men often do not correctly understand their own individual behaviour or its sources; to stop searching for causes, in the most literal and mechanical sense, in accounting for what happened and how - all this would be absurdly childish and frivolous (not to say obscurantist), and I did not suggest anything of this kind. But to ignore motives and the context in which they arose, the range of possibilities as they stretched before the actors, most of which never were, and some never could have been, realised; to ignore the spectrum of human thought and imagination - how the world and they themselves appear to men whose vision and values (illusions and all) we can grasp in the end only in terms of our own - would be to cease to write history. One may argue about the degree of difference that the influence of this or that individual made in shaping events. But to try to reduce the behaviour of individuals to that of impersonal 'social forces' not further analysable into the conduct of the men who, even according to Marx, make history is 'reification' of statistics, a form of the 'false consciousness' of bureaucrats and administrators who close their eyes to all that proves incapable of quantification, and thereby perpetrate absurd­ities in theory and dehumanisation in practice.

There are remedies that breed new diseases, whether or not they cure those to which they are applied. To frighten human beings by suggesting to them that they are in the grip of impersonal forces over which they have little or no control is to breed myths, ostensibly in order to kill other figments - the notion of superna­tural forces, or of all-powerful individuals, or of the invisible hand. It is to invent entities, to propagate faith in unalterable patterns of events for which the empirical evidence is, to say the least, insufficient, and which by relieving individuals of the burdens of personal responsibility breeds irrational passivity in some, and no less irrational fanatical activity in others; for nothing is more inspiring than the certainty that the stars in their courses are fighting for one's cause, that 'History', or 'social forces', or 'the wave of the future' are with one, bearing one aloft and forward.

This way of thinking and speaking is one which it is the great merit of modern empiricism to have exposed. If my essay has any p_olemical thrust, it is to discredit metaphysical constructions of this kind. If to speak of men solely in terms of statistical probabilities, ignoring too much of what is specificalJy human in men - evaluations, choices, differing visions of life - is an exaggerated application of scientific method, a gratuitous behav­iourism, it is no less misleading to appeal to imaginary forces. The former has its place; it describes, classifies, predicts, even if it does not explain. The latter explains indeed, but in occult, what I can only call neo-animistic, terms. I suspect that Carr does not feel anxious to defend either of these methods. But in his reaction against naivete, smugness, the vanities of nationalistic or class or personal moralising, he has permitted himself to be driven to the other extreme - the night of impersonality, in which human beings are dissolved into abstract forces. The fact that I protest against it leads Carr to think that I embrace the opposite absurdity. His assumption that between them these extremes exhaust the possibil­ities seems to me to be the basic fallacy from which his (and perhaps others') vehement criticism of my real and imaginary opinions ultimately stems.

At this point I should like to reiterate some commonplaces from which I do not depart: that causal laws are applicable to human history (a proposition which, pace Carr, I should consider it insane to deny); that history is not mainly a 'dramatic conflict' between individual wills;¹ that knowledge, especially of scientifically estab­lished laws, tends to render us more effective[21] and extend our liberty, which is liable to be curtailed by ignorance and the illusions, terrors and prejudices that it breeds;[22] that there is plenty of empirical evidence for the view that the frontiers of free choice are a good deal narrower than many men have in the past supposed, and perhaps still erroneously believe;² and even that objective patterns in history may, for all I know, be discernible. I must repeat that my concern is only to assert that unless such laws and patterns are held to leave some freedom of choice - and not only freedom of action determined by choices that are themselves wholly determined by antecedent causes - we shall have to reconstruct our view of reality accordingly; and that this task is far more formidable than determinists tend to assume.

The determinist's world may, at least in principle, be conceiv­able: in it all that Ernest Nagel declares to be the function of human volition will remain intact; a man's behaviour will still be affected by praise and blame as his metabolism (at any rate directly) will not;³ men will continue to describe persons and things as beautiful or ugly, evaluate actions as beneficial or harmful, brave or cowardly, noble or ignoble. But when Kant said that if the laws that governed the phenomena of the external world turned out to govern all there was, then morality - in his sense - was annihilated; and when, in consequence, being concerned with the concept of freedom presupposed by his notion of moral responsibility, he adopted very drastic measures in order to save it, he seems to me, at the very least, to have shown a profound grasp of what is at stake. His solution is obscure, and perhaps untenable; but although it may have to be rejected, the problem remains. In a causally determined system the notions of free choice and moral responsi­bility, in their usual senses, vanish, or at least lack application, and the notion of action would have to be reconsidered.

I recognise the fact that some thinkers seem to feel no intellectual discomfort in interpreting such concepts as responsibility, culpabil­ity, remorse in strict conformity with causal determination. At most they seek to explain the resistance of those who dissent from them by attributing to them a confusion of causality with some sort of compulsion. Compulsion frustrates my wishes but when I fulfil them I am surely free, even though my wishes themselves are causally determined; if they are not, if they are not effects of my general tendencies, or ingredients in my habits and way of living (which can be described in purely causal terms), or if these, in their turn, are not what they are entirely as a result of causes - physical, social, psychological - then there is surely an element of pure chance or randomness, which breaks the causal chain. But (the argument continues) is not random behaviour the very opposite of freedom, rationality, responsibility? And yet these alternatives seem to exhaust the possibilities. The notion of uncaused choice as something out of the blue is certainly not satisfactory. But (I need not argue this again) the only alternative permitted by such thinkers - a caused choice held to entail responsibility, desert and the like - is equally untenable.

This dilemma has now divided thinkers for more than two thousand years. Some continue to be agonised or at least puzzled by it, as the earliest Stoics were; others see no problem at all. It may be that it stems, at least in part, from the use of a mechanical model applied to human actions; in one case choices are conceived as links in the kind of causal sequence that is typical of the functioning of a mechanical process; in the other, a break in this sequence, still conceived in terms of a highly complex mechanism. Neither image seems to fit the case at all well. We seem to need a new model, a schema which will rescue the evidence of moral consciousness from the beds of Procrustes provided by the obsessive frameworks of the traditional discussions. All efforts to break away from the old obstructive analogies, or (to use a familiar terminology) the rules of an inappropriate language game, have so far proved abortive. This needs a philosophical imagination of the first order, which in this case is still to seek. White's solution - to attribute the conflicting views to different scales of value or varieties of moral usage - seems to me no way out. I cannot help suspecting that his view is part of a wider theory, according to which belief in determinism or any other view of the world is, or depends on, some sort of large-scale pragmatic decision about how to treat this or that field of thought or experience, based on a view of what set of categories would give the best results. Even if one accepted this, it would not be easier to reconcile the notions of causal necessity, avoidability, free choice, responsibility and the rest. I do not claim to have refuted the conclusions of determinism; but neither do J see why we need be driven towards them. Neither the idea of historical explanation as such, nor respect for scientific method, seems to me to entail them.

This sums up my disagreements with Ernest Nagel, Morton White, E. H. Carr, the classical determinists and their modern disciples.

ii

Positive versus negative liberty

In the case of social and political libeny a problem arises that is not wholly dissimilar from that of social and historical determinism. We assume the need of an area for free choice, the diminution of which is incompatible with the existence of anything that can properly be called political (or social) liberty. Indeterminism does not entail that human beings cannot in fact be treated like animals or things; nor is political liberty, like freedom of choice, intrinsic to the notion of a human being; it is a historical growth, an area bounded by frontiers. The question of its frontiers, indeed whether the concept of frontiers can properly be applied to it, raises issues on which much of the criticism directed upon my theses has concentrated. The main issues may here too be summarised under three heads: (a) whether the difference I have drawn between (what I am not the first to have called) positive and negative liberty is specious, or, at any rate, too sharp; (b) whether the term 'liberty' can be extended as widely as some of my critics appear to wish, without thereby depriving it of so much significance as to render it progressively less useful; (c) why political liberty should be regarded as being of value.

Before discussing these problems, I wish to correct a genuine error in the original version of Two Concepts of Liberty. Although this error does not weaken, or conflict with, the arguments used in the essay (indeed, if anything, it seems to me to strengthen them), it is, nevertheless, a position that I consider to be mistaken.[23] In the original version of Two Concepts of Liberty[24] I speak of liberty as the absence of obstacles to the fulfilment of a man's desires. This is a common, perhaps the most common, sense in which the term is used, but it does not represent my position. For if to be free - _negatively - is simply not to be prevented by other persons from doing whatever one wishes, then one of the ways of attaining such freedom is by extinguishing one's wishes. I offered criticisms of this definition, and of this entire line of thought, in the text, without realising that it was inconsistent with the formulation with which 1 began. If degrees of freedom were a function of the satisfaction of desires, I could increase freedom as effectively by eliminating desires as by satisfying them: I could render men (including myself) free by conditioning them into losing the original desires which I have decided not to satisfy. Instead of resisting or removing the pressures that bear down upon me, I can 'internalise' them. This is what Epictetus achieves when he claims that he, a slave, is freer than his master. By ignoring obstacles, forgetting, 'rising above' them, becoming unconscious of them, 1 can attain peace and serenity, a noble detachment from the fears and hatreds that beset other men - freedom in one sense indeed, but not in the sense in which I wish to speak of it. When (according to Cicero's account) the Stoic sage Posidonius, who was dying of an agonising disease, said, 'Do your worst, pain; no matter what you do, you cannot make me hate you',[25] thereby accepting, and attaining unity with, 'Nature', which, being identical with cosmic 'reason', rendered his pain not merely inevitable, but rational, the sense in which he achieved freedom is not that basic meaning of it in which men are said to lose freedom when they are imprisoned or literally enslaved. The Stoic sense of freedom, however sublime, must be distinguished from the freedom or liberty which the oppressor, or the oppressive institutionalised practice, curtails or destroys. [26] For once I am happy to acknow­ledge the insight of Rousseau: to know one's chains for what they are is better than to deck them with flowers.¹

Spiritual freedom, like moral victory, must be distinguished from a more fundamental sense of freedom, and a more ordinary sense of victory, otherwise there will be a danger of confusion in theory and justification of oppression in practice, in the name of liberty itself. There is a clear sense in which to teach a man that, if he cannot get what he wants, he must learn to want only what he can get, may contribute to his happiness or his security; but it will not increase his civil or political freedom. The sense of freedom in which I use this term entails not simply the absence of frustration (which may be obtained by killing desires), but the absence of obstacles to possible choices and activities - absence of obstruc­tions on roads along which a man can decide to walk. Such freedom ultimately depends not on whether I wish to walk at all, or how far, but on how many doors are open, how open they are, upon their relative importance in my life, even though it may be impossible literally to measure this in any quantitative fashion.² The extent of my social or political freedom consists in the absence of obstacles not merely to my actual, but to my potential, choices - to my acting in this or that way if I choose to do so. Similarly absence of such freedom is due to the closing of such doors or failure to open them, as a result, intended or unintended, of alterable human practices, of the operation of human agencies; although only if such acts are deliberately intended (or, perhaps, are accompanied by awareness that they may block paths) will they be liable to be called oppression. Unless this is conceded, the Stoic conception of liberty ('true' freedom - the state of the morally autonomous slave), which is compatible with a very high degree of political despotism, will merely confuse the issue.

It is an interesting, but perhaps irrelevant, historical question at

which we have become all too familiar. There is, of course, all the difference in the world between assimilating the rules of reason, as advocated by Stoicism, and those of an irrational movement or arbitrary dictatorship. But the psychological machinery is similar.

' This point is well made by one of my critics, L. J. Macfarlane, in 'On Two Concepts of Liberty', Political Studies 14 (1966), 77-81. In the course of a very critical but fair and valuable discussion Macfarlane observes that to know one's chains is often the first step to freedom, which may never come about if one either ignores or loves them.

² See p. i 77 below, note 1.

what date, and in what circumstances, the notion of individual liberty in this sense first became explicit in the West. I have found _no convincing evidence of any clear formulation of it in the ancient world.¹ Some of my critics have doubted this, but apart from pointing to such modern writers as Acton, Jellinek or Barker, who do profess to find this ideal in ancient Greece, some of them also, more pertinently, cite the proposals of Otanes after the death of pseudo-Smerdis in the account given by Herodotus, and the celebrated paean to liberty in the Funeral Oration of Pericles, as well as the speech of Nikias before the final battle with the Syracusans (in Thucydides), as evidence that the Greeks, at any rate, had a clear conception of individual liberty. I must confess that I do not find this conclusive. When Pericles and Nikias compare the freedom of the Athenian citizens with the fate of the subjects of less democratic States, what (it seems to me) they are saying is that the citizens of Athens enjoy freedom in the sense of self-government, that they are not slaves of any master, that they perform their civic duties out of love for their polls, without needing to be coerced, and not under the goads and whips of savage laws or taskmasters (as in Sparta or Persia). So might a headmaster say of the boys in his school that they live and act according to good principles not because they are forced to do so, but because they are inspired by loyalty to the school, by 'team spirit', by a sense of solidarity and common purpose; whereas at other schools these results have to be achieved by fear of punishment and stern measures. But in neither case is it contem­plated that a man might, without losing face, or incurring contempt, or a diminution of his human essence, withdraw from public life altogether, and pursue private ends, live in a room of his own, in the company of personal friends, as Epicurus later advocated, and perhaps the Cynic and Cyrenaic disciples of Socrates had preached before him. As for Otanes, he wished neither to rule nor to be ruled - the exact opposite of Aristotle's notion of true civic liberty. Perhaps this attitude did begin to occur in the ideas of unpolitical thinkers in Herodotus' day: of Antiphon the Sophist, for example, and possibly in some moods of Socrates himself. But it remains isolated and, until Epicurus, undeveloped. In other words, it seems to me that the issue of individual freedom,

' For a fuller treatment by Berlin of liberty in the ancient world see now 'The Birth of Greek Individualism', reprinted below. Ed.

of the frontiers beyond which public authority, whether lay or ecclesiastical, should not normally be allowed to step, had not clearly emerged at this stage; the central value attached to it may, perhaps (as I remarked in the penultimate paragraph of my lecture), be the late product of a capitalist civilisation, an element in a network of values that includes such notions as personal rights, civil liberties, the sanctity of the individual personality, the importance of privacy, personal relations and the like. I do not say that the ancient Greeks did not in fact enjoy a great measure of what we should today call individual liberty.[27] My thesis is only that the notion had not explicitly emerged, and was therefore not central to Greek culture, or, perhaps, to any other ancient civilisation known to us.

One of the by-products or symptoms of this stage of social development is that, for instance, the issue of free will (as opposed to that of voluntary action) is not felt to be a problem before the Stoics; the corollary of which seems to be that variety for its own sake - and the corresponding abhorrence of uniformity - is not a prominent ideal, or perhaps an explicit ideal at all, before the Renaissance, or even, in its full form, before the beginning of the eighteenth century. Issues of this type seem to arise only when forms of life, and the social patterns that are part of them, after long periods in which they have been taken for granted, are upset, and so come to be recognised and become the subject of conscious reflection. There are many values which men have disputed, and for and against which they have fought, that are not mentioned in some earlier phase of history, either because they are assumed without question, or because men are, whatever the cause, in no condition to conceive of them. It may be that the more sophisti­cated forms of individual liberty did not impinge upon the consciousness of the masses of mankind simply because they lived in squalor and oppression. Men who live in conditions where there is not sufficient food, warmth, shelter, or the minimum degree of security, can scarcely be expected to concern themselves with freedom of contract or of the press.

It may make matters clearer if at this point I mention what seems to me yet another misconception - namely the identification of freedom with activity as such. When, for example, Erich Fromm, in virtually all his moving tracts for the times, speaks of true freedom as the spontaneous, rational activity of the total, integrated personality, and is partly followed in this by Bernard Crick,[28] I disagree with them. The freedom of which I speak is opportunity for action, rather than action itself. If, although I enjoy the right to walk through open doors, I prefer not to do so, but to sit still and vegetate, I am not thereby rendered less free. Freedom is the opportunity to act, not the action itself; the possibility of action, not necessarily that dynamic realisation of it which both Fromm and Crick identify with it. If apathetic neglect of various avenues to a more vigorous and generous life - however much this may be condemned on other grounds - is not considered incompatible with the notion of being free, then I have nothing to quarrel with in the formulations of either of these writers. But I fear that Fromm would consider such abdication as a symptom of lack of integration, which for him is indispensable to - perhaps identical with - freedom; while Crick would look upon such apathy as too inert and timid to deserve to be called freedom. I find the ideal advocated by these champions of the full life sympathetic; but to identify it with freedom seems to me conflation of two values. To say that freedom is activity as such is to make the term cover too much; it tends to obscure and dilute the central issue - the right and freedom to act - about which men have argued and fought during almost the whole of recorded history.

To return to concepts of liberty. Much has been made by my opponents of the distinction (regarded by them as specious or exaggerated) that I have tried to draw between two questions: 'By whom am I governed?' and 'How much am I governed?' Yet I confess that I cannot see either that the two questions are identical, or that the difference between them is unimportant. It still seems to me that the distinction between the two kinds of answer, and therefore between the different senses of 'liberty' involved, is neither trivial nor confused. Indeed, I continue to believe that the issue is a central one both historically and conceptually, both in theory and practice.

Let me say once again that 'positive' and 'negative' liberty, in the sense in which I use these terms, start at no great logical distance from each other. The questions 'Who is master?' and 'Over what area am I master?' cannot be kept wholly distinct. I wish to determine myself, and not be directed by others, no matter how wise and benevolent; my conduct derives an irreplaceable value from the sole fact that it is my own, and not imposed upon me. But I am not, and cannot expect to be, wholly self-sufficient or socially omnipotent.[29] I cannot remove all the obstacles in my path that stem from the conduct of my fellows. I can try to ignore them, treat them as illusory, or 'intermingle' them and attribute them to my own inner principles, conscience, moral sense; or try to dissolve my sense of personal identity in a common enterprise, as an element in a larger self-directed whole. Nevertheless, despite such heroic efforts to transcend or dissolve the conflicts and resistance of others, if I do not wish to be deceived, I shall recognise the fact that total harmony with others is incompatible with self-identity; that if I am not to be dependent on others in every respect, I shall need some area within which I am not, and can count on not being, freely interfered with by them. The question then arises: how wide is the area over which I am, or should be, master? My thesis is that historically the notion of 'positive' liberty - in answer to the question 'Who is master?' - diverged from that of 'negative' liberty, designed to answer 'Over what area am I master?'; and that this gulf widened as the notion of the self suffered a metaphysical fission into, on the one hand, a 'higher', or a 'real', or an 'ideal' self, set up to rule a 'lower', 'empirical', 'psychological' self or nature, on the other; into 'myself at my best' as master over my inferior day-to-day self; into Coleridge's great I am over less transcendent incarnations of it in time and space.[30]

A genuine experience of inner tension may lie at the root of this ancient and pervasive metaphysical image of the two selves, the influence of which has been vast over language, thought and conduct; however this may be, the 'higher' self duly became identified with institutions, Churches, nations, races, States, classes, cultures, parties, and with vaguer entities, such as the general will, the common good, the enlightened forces of society, the vanguard of the most progressive class, Manifest Destiny. My thesis is that, in the course of this process, what had begun as a doctrine of freedom turned into a doctrine of authority and, at times, of oppression, and became the favoured weapon of despotism, a phenomenon all too familiar in our own day. I was careful to point out that this could equally have been the fate of the doctrine of negative liberty. Among the dualists who distinguished the two selves, some - in particular Jewish and Christian theologians, but also Idealist metaphysicians in the nineteenth century - speak of the need to release the 'higher' or 'ideal' self from obstacles in its path, such as interference by, 'slavery to', the 'lower' self; and some saw this base entity incarnated in institutions serving irrational or wicked passions and other forces of evil likely to obstruct the proper development of the 'true' or 'higher' self, or 'myself at my best'. The history of political doctrines might (like that of some Protestant sects) have taken this 'negative' form. The point, however, is that it did so relatively seldom - as, for example, in early liberal, anarchist and some types of populist writings. But for the most part freedom was identified, by metaphysically inclined writers, with the realisation of the real self not so much in individual men as incarnated in institutions, traditions, forms of life wider than the empirical spatio-temporal existence of the finite individual. Freedom is identified by such thinkers most often, it seems to me, with the 'positive' activity of these institutional ('organic') forms of life, growth and so forth rather than with mere ('negative') removal of obstacles even from the paths of such 'organisms', let alone from those of individuals - such an absence of obstacles being regarded as, at best, a means to, or a condition of, freedom; not as freedom itself.

It is doubtless well to remember that belief in negative freedom ^Is compatible with, and (so far as ideas influence conduct) has played its part in generating, great and lasting social evils. My point i^S that it was much less often defended or disguised by the kind of specious arguments and sleights-of-hand habitually used by the champions of 'positive' freedom in its more sinister forms. Advocacy of non-interference (like 'social Darwinism') was, of course, used to support politically and socially destructive policies which armed the strong, the brutal and the unscrupulous against the humane and the weak, the able and ruthless against the less gifted and the less fortunate. Freedom for the wolves has often meant death to the sheep. The bloodstained story of economic individualism and unrestrained capitalist competition does not, I should have thought, today need stressing. Nevertheless, in view of the astonishing opinions which some of my critics have imputed to me, I should, perhaps, have been wise to underline certain parts of my argument. I should have made even clearer that the evils of unrestricted laissez-faire, and of the social and legal systems that permitted and encouraged it, led to brutal violations of 'negative' liberty - of basic human rights (always a 'negative' notion: a wall against oppressors), including that of free expression or association, without which there may exist justice and fraternity and even happiness of a kind, but not democracy. And I should perhaps have stressed (save that I thought this too obvious to need saying) the failure of such systems to provide the minimum conditions in which any degree of significant 'negative' liberty can be exercised by individuals or groups, and without which it is of little or no value to those who may theoretically possess it. For what are rights without the power to implement them?

I had supposed that enough had been said by almost every serious modern writer concerned with this subject about the fate of personal liberty during the reign of unfettered economic individu­alism - about the condition of the injured majority, principally in the towns whose children were destroyed in mines or mills while their parents lived in poverty, disease and ignorance, a situation in which the enjoyment by the poor and the weak of legal rights to spend their money as they pleased or to choose the education they wanted (which Cobden and Herbert Spencer and their disciples offered them with every appearance of sincerity) became an odious mockery.

All this is notoriously true. Legal liberties are compatible with extremes of exploitation, brutality and injustice. The case for intervention, by the State or other effective agencies, to secure conditions for both positive, and at least a minimum degree of negative, liberty for individuals, is overwhelmingly strong. Liberals like Tocqueville and J. S. Mill, and even Benjamin Constant (who prized negative liberty beyond any modern writer) were not unaware of this. The case for social legislation or planning, for the

Welfare State and socialism, can be constructed with as much validity from consideration of the claims of negative liberty as from those of its positive brother, and if, historically, it was not made so frequently, that was because the kind of evil against which the concept of negative liberty was directed as a weapon was not laissez-faire, but despotism. The rise and fall of the two concepts can largely be traced to the specific dangers which, at a given _moment, threatened a group or society most: on the one hand excessive control and interference, or, on the other, an uncon­trolled 'market' economy. Each concept seems liable to perversion into the very vice which it was created to resist. But whereas liberal ultra-individualism could scarcely be said to be a rising force at present, the rhetoric of 'positive' liberty, at least in its distorted form, is in far greater evidence, and continues to play its historic role (in both capitalist and anti-capitalist societies) as a cloak for despotism in the name of a wider freedom.

'Positive' liberty, conceived as the answer to the question, 'By whom am I to be governed?', is a valid universal goal. I do not know why I should have been held to doubt this, or, for that matter, the further proposition, that democratic self-government is a fundamental human need, something valuable in itself, whether or not it clashes with the claims of negative liberty or of any other goal; valuable intrinsically and not only for the reasons advanced in its favour by, for example, Constant - that without it negative liberty may be too easily crushed; or by Mill, who thinks it an indispensable means - but still only a means - to the attainment of happiness. I can only repeat that the perversion of the notion of positive liberty into its opposite - the apotheosis of authority - did occur, and has for a long while been one of the most familiar and depressing phenomena of our time. For whatever reason or cause, the notion of 'negative' liberty (conceived as the answer to the question 'How much am I to be governed?'), however disastrous the consequences of its unbridled forms, has not historically been twisted by its theorists as often or as effectively into anything so darkly metaphysical or socially sinister or remote from its original meaning as its 'positive' counterpart. The first can be turned into its opposite and still exploit the favourable associations of its innocent origins. The second has, much more frequently, been seen, for better and for worse, for what it was; there has been no lack of emphasis, in the last hundred years, upon its more disastrous implications. Hence the greater need, it seems to me, to expose the aberrations of positive liberty than those of its negative brother.

Nor do I wish to deny that new ways in which liberty, in both its positive and its negative sense, can be, and has been, curtailed have arisen since the nineteenth century. In an age of expanding economic productivity there exist ways of curtailing both types of liberty - for example, by permitting or promoting a situation in which entire groups and nations are progressively shut off from benefits which have been allowed to accumulate too exclusively in the hands of other groups and nations, the rich and strong - a situation which, in its turn, has produced (and was itself produced by) social arrangements that have caused walls to arise around men, and doors to be shut to the development of individuals and classes. This has been done by social and economic policies that were sometimes openly discriminatory, at other times camouflaged, by the rigging of educational policies and of the means of influencing opinions, by legislation in the sphere of morals, and by similar measures, which have blocked and diminished human freedom at times as effectively as the more overt and brutal methods of direct oppression - slavery and imprisonment - against which the original defenders of liberty lifted their voices.[31]

Let me summarise my position thus far. The extent of a man's _neg_ative freedom is, as it were, a function of what doors, and how many, are open to him; upon what prospects they open; and how open they are. This formula must not be pressed too far, for not all doors are of equal importance, inasmuch as the paths on which they open vary in the opportunities they offer. Consequently, the problem of how an overall increase of liberty in particular circumstances is to be secured, and how it is to be distributed (especially in situations, and this is almost invariably the case, in which the opening of one door leads to the lifting of other barriers and the lowering of still others), how, in a word, the maximisation of opportunities is in any concrete case to be achieved, can be an agonising problem, not to be solved by any hard-and-fast rule.¹

freedom, and of the no less fatal identification of conditions of freedom with freedom itself, which is at the root of some of the fallacies with which I am concerned.

¹ David Nicholls in an admirable survey, 'Positive Liberty, 188(^1914', American Political Science Review 56 (1962), I 14-28, at I 14 note 8, thinks that I contradict myself in quoting with approval Bentham's view that every law is an infraction of liberty (see p. 195 below, note 1), since some laws increase the total amount of liberty in a society. I do not see the force of this objection. Every law seems to me to curtail some liberty, although it may be a means to increasing another. Whether it increases the total sum of attainable liberty will of course depend on the particular situation. Even a law which enacts that no one shall coerce anyone in a given sphere, while it obviously increases the freedom of the majority, is an infraction of the freedom of potential bullies and policemen. Infraction may, as in this case, be highly desirable, but it remains infraction. There is no reason for thinking that Bentham, who favoured laws, meant to say more than this.

In his article (at p. 121, note 63) Nicholls quotes T. H. Green's statement (in his 'Lecture on "Liberal Legislation and Freedom of Contract"'): 'the mere removal of compulsion, the mere enabling a man to do as he likes, is in itself no contribution to true freedom . . . the ideal of true freedom is the maximum of power for all members of human society alike to make the best of themselves': pp. 199-200 in T. H. Green, Lectures on the Principles of Political Obligation and Other Writings, ed. Paul Harris and John Morrow (Cambridge etc., 1986). This is a classical statement of positive liberty, and the crucial terms are, of course, 'true freedom' and 'the best of themselves'. Perhaps I need not enlarge again upon the fatal ambiguity of these words. As a plea for justice, and a denunciation of the monstrous assumption that workmen were (in any sense that mattered to them) ^free agents in negotiating with employers in his time, Green's essay can scarcely be improved upon. The workers, in theory, probably enjoyed wide negative ^freedom. But since they lacked the means of its realisation, it was a hollow gain. Hence I find nothing to disagree with in Green's recommendations; I reject only the metaphysical doctrine of the two selves - the individual streams versus the

What I am mainly concerned to establish is that, whatever may be the common ground between them, and whichever is liable to graver distortion, negative and positive liberty are not the same thing. Both are ends in themselves. These ends may clash irreconcilably. When this happens, questions of choice and preference inevitably arise. Should democracy in a given situation be promoted at the expense of individual freedom? Or equality at the expense of artistic achievement; or mercy at the expense of justice; or spontaneity at the expense of efficiency; or happiness, loyalty, innocence at the expense of knowledge and truth? The simple point which I am concerned to make is that where ultimate values are irreconcilable, clear-cut solutions cannot, in principle, be found. To decide rationally in such situations is to decide in the light of general ideals, the overall pattern of life pursued by a man or a group or a society. If the claims of two (or more than two) types of liberty prove incompatible in a particular case, and if this is an instance of the clash of values at once absolute and incommensurable, it is better to face this intellectually uncomfort­able fact than to ignore it, or automatically attribute it to some deficiency on our part which could be eliminated by an increase in skill or knowledge, as was done by Condorcet and his disciples; or, what is worse still, suppress one of the competing values altogether by pretending that it is identical with its rival - and so end by distorting both. Yet, it appears to me, it is exactly this that philosophical monists who demand final solutions - tidiness and harmony at any price - have done and are doing still. I do not, of course, mean this as an argument against the proposition that the application of knowledge and skill can, in particular cases, lead to satisfactory solutions. When such dilemmas arise it is one thing to say that every effort must be made to resolve them, and another that it is certain a priori that a correct, conclusive solution must always in principle be discoverable - something that the older rationalist metaphysics appeared to guarantee.

social river in which they should be merged, a dualistic fallacy used too often to support a variety of despotisms. Nor, of course, do I wish to deny that Green's views were exceptionally enlightened; and this holds of many of the critics of liberalism in Europe and America in the last hundred years or so. Nevertheless, words are important, and a writer's opinions and purposes are not sufficient to render the usc of a misleading terminology harmless either in theory or in practice. The record of liberalism is no better in this respect that that of most other schools of political thought.

Consequently, when another of my critics, David Spitz,¹ _maintains that the frontier falls not so much between positive and _negative liberty, but 'in determination of which complex of particular liberties and concomitant restraints is most likely to promote those values that, in Berlin's theory, are distinctively human', and, in the course of his interesting and suggestive review, declares that the issue depends on one's view of human nature, or of human goals (on which men differ), I do not dissent. But when he goes on to say that, in my attempt to cope with the relativity of values, I fall back on the views of J. S. Mill, he seems to me mistaken on an important issue. Mill does seem to have convinced himself that there exists such a thing as attainable, communicable, objective truth in the field of value judgements; but that the conditions for its discovery do not exist save in a society which provides a sufficient degree of individual liberty, particularly of enquiry and discussion. This is simply the old objectivist thesis, in an empirical form, with a special rider about the need for individual liberty as a necessary condition for the attainment of this final goal. My thesis is not this at all; but that, since some values may conflict intrinsically, the very notion that a pattern must in principle be discoverable in which they are all rendered harmonious is founded on a false a priori view of what the world is like.

If I am right in this, and the human condition is such that men cannot always avoid choices, they cannot avoid them not merely for the obvious reasons, which philosophers have seldom ignored, namely that there are many possible courses of action and forms of life worth living, and therefore to choose between them is part of being rational or capable of moral judgement; they cannot avoid choice for one central reason (which is, in the ordinary sense, conceptual, not empirical), namely that ends collide; that one cannot have everything. Whence it follows that the very concept of an ideal life, a life in which nothing of value need ever be lost or sacrificed, in which all rational (or virtuous, or otherwise legiti­mate) wishes must be capable of being truly satisfied - this classical vision is not merely Utopian, but incoherent. The need to choose, ^to sacrifice some ultimate values to others, turns out to be a ^permanent characteristic of the human predicament. If this is so, it ^undermines all theories according to which the value of free choice ^derives from the fact that without it we cannot attain to the perfect

' op. cit. (p. xiii above, note 2).

life; with the implication that once such perfection has been reached the need for choice between alternatives withers away. On this view, choice, like the party system, or the right to vote against the nominees of the ruling party, becomes obsolete in the perfect Platonic or theocratic or Jacobin or communist society, where any sign of the recrudescence of basic disagreement is a symptom of error and vice. For there is only one possible path for the perfectly rational man, since there are now no beguiling illusions, no conflicts, no incongruities, no surprises, no genuine, unpredictable novelty; everything is still and perfect in the universe governed by what Kant called the Holy Will.

Whether or not this calm and tideless sea is conceivable or not, it does not resemble the real world in terms of which alone we conceive men's nature and their values. Given things as we know them, and have known them during recorded human history, capacity for choosing is intrinsic to rationality, if rationality entails a normal ability to apprehend the real world. To move in a frictionless medium, desiring only what one can attain, not tempted by alternatives, never seeking incompatible ends, is to live in a coherent fantasy. To offer it as the ideal is to seek to dehumanise men, to turn them into the brainwashed, contented beings of Aldous Huxley's celebrated totalitarian nightmare. To contract the areas of human choice is to do harm to men in an intrinsic, Kantian, not merely utilitarian, sense. The fact that the maintenance of conditions making possible the widest choice must be adjusted - however imperfectly - to other needs, for social stability, predictability, order and so on - does not diminish their central importance. There is a minimum level of opportunity for choice - not of rational or virtuous choice alone - below which human activity ceases to be free in any meaningful sense. It is true that the cry for individual liberty has often disguised desire for privilege, or for power to oppress and exploit, or simply fear of social change. Nevertheless the modern horror of uniformity, conformism and mechanisation of life is not groundless.

As for the issue of relativity and the subjective nature of values, I wonder whether this has not, for the sake of argument, been exaggerated by philosophers: whether men and their outlooks have differed, over wide stretches of space and time, as greatly as has at times been represented. But on this point - how unchanging, how 'ultimate', how universal and 'basic' human values are - I feel no certainty. If values had varied very widely between cultures and periods, communication would have been harder to achieve, and our historical knowledge, which depends on some degree of ability to understand the goals and motives and ways of life at work in cultures different from our own, would turn out to be an illusion. So, of course, would the findings of historical sociology, from which the very concept of social relativity largely derives. Scepti­cism, driven to extremes, defeats itself by becoming self-refuting.

As for the question of what in fact are the values which we regard as universal and 'basic' - presupposed (if that is the correct logical relation) by the very notions of morality and humanity as _such - this seems to me a question of a quasi-empirical kind. That is to say, it seems to be a question for the answer to which we must go to historians, anthropologists, philosophers of culture, social scientists of various kinds, scholars who study the central notions and central ways of behaviour of entire societies, revealed in monuments, forms of life, social activity, as well as more overt expressions of belief such as laws, faiths, philosophies, literature. I describe this as quasi-empirical, because concepts and categories that dominate life and thought over a very large portion (even if not the whole) of recorded history are difficult, and in practice impossible, to think away; and in this way differ from the more flexible and changing constructions and hypotheses of the natural sciences.

There is one further point which may be worth reiterating. It is important to discriminate between liberty and the conditions of its exercise. If a man is too poor or too ignorant or too feeble to make use of his legal rights, the liberty that these rights confer upon him is nothing to him, but it is not thereby annihilated. The obligation to promote education, health, justice, to raise standards of living, to provide opportunity for the growth of the arts and the sciences, to prevent reactionary political or social or legal policies or arbitrary inequalities, is not made less stringent because it is not necessarily directed to the promotion of liberty itself, but to conditions in which alone its possession is of value, or to values which may be independent of it. And still, liberty is one thing, and the conditions for it are another. To take a concrete example: it is, I believe, desirable to introduce a uniform system of general primary and secondary education in every country, if only in order to do away with distinctions of social status that are at present created or promoted by the existence of a social hierarchy of schools in some

Western countries, notably my own. If I were asked why I believe this, I should give the kind of reasons mentioned by Spitz,¹ for instance, the intrinsic claims of social equality; the evils arising from differences of status created by a system of education governed by the financial resources or the social position of parents rather than the ability and the needs of the children; the ideal of social solidarity; the need to provide for the bodies and minds of as many human beings as possible, and not only of members of a privileged class; and, what is more relevant here, the need to provide the maximum number of children with opportunities for free choice, which equality in education is likely to increase.

If I were told that this must severely curtail the liberty of parents who claim the right not to be interfered with in this matter - that it was an elementary right to be allowed to choose the type of education to be given to one's child, to determine the intellectual, religious, social, economic conditions in which the child is to be brought up - I should not be ready to dismiss this outright. But I should maintain that when (as in this case) values genuinely clash, choices must be made. In this case the clash arises between the need to preserve the existing liberty of some parents to determine the type of education they seek for their children; the need to promote other social purposes; and, finally, the need to create conditions in which those who lack them will be provided with opportunities to exercise those rights (freedom to choose) which they legally possess, but cannot, without such opportunities, put to use. Useless freedoms should be made usable, but they are not identical with the conditions indispensable for their utility. This is not a merely pedantic distinction, for if it is ignored, the meaning and value of freedom of choice is apt to be downgraded. In their zeal to create social and economic conditions in which alone freedom is of genuine value, men tend to forget freedom itself; and if it is remembered, it is liable to be pushed aside to make room for these other values with which the reformers or revolutionaries have become preoccupied.

Again, it must not be forgotten that even though freedom without sufficient material security, health, knowledge, in a society that lacks equality, justice, mutual confidence, may be virtually useless, the reverse can also be disastrous. To provide for material

' ibid., p. So.

needs, for education, for such equality and security as, say, children have at school or laymen have in a theocracy, is not to expand liberty. We live in a world characterised by regimes (both right- and left-wing) which have done, or are seeking to do, precisely this; and when they call it freedom, this can be as great a fraud as the freedom of the pauper who has a legal right to purchase luxuries. Indeed, one of the things that Dostoevsky's celebrated fable of the Grand Inquisitor in The Brothers Karamazov is designed to show is precisely that paternalism can provide the conditions of freedom, yet withhold freedom itself.

A general consideration follows. If we wish to live in the light of reason, we must follow rules or principles; for that is what being rational is. When these rules or principles conflict in concrete cases, to be rational is to follow the course of conduct which least obstructs the general pattern of life in which we believe. The right policy cannot be arrived at in a mechanical or deductive fashion: there are no hard-and-fast rules to guide us; conditions are often unclear, and principles incapable of being fully analysed or articulated. We seek to adjust the unadjustable, we do the best we can. Those, no doubt, are in some way fortunate who have brought themselves, or have been brought by others, to obey some ultimate principle before the bar of which all problems can be brought. Single-minded monists, ruthless fanatics, men possessed by an all- embracing coherent vision do not know the doubts and agonies of those who cannot wholly blind themselves to reality. But even those who are aware of the complex texture of experience, of what is not reducible to generalisation or capable of computation, can, in the end, justify their decisions only by their coherence with some overall pattern of a desirable form of personal or social life, of which they may become fully conscious only, it may be, when faced with the need to resolve conflicts of this kind. If this seems vague, it is so of necessity. The notion that there must exist final objective answers to normative questions, truths that can be ^demonstrated or directly intuited, that it is in principle possible to ^discover a harmonious pattern in which all values are reconciled, ^and that it is towards this unique goal that we must make; that we can uncover some single central principle that shapes this vision, a ^principle which, once found, will govern our lives - this ancient and almost universal belief, on which so much traditional thought and action and philosophical doctrine rests, seems to me invalid, and at times to have led (and still to lead) to absurdities in theory and barbarous consequences in practice.[32]

The fundamental sense of freedom is freedom from chains, from imprisonment, from enslavement by others. The rest is extension of this sense, or else metaphor. To strive to be free is to seek to remove obstacles; to struggle for personal freedom is to seek to curb interference, exploitation, enslavement by men whose ends are theirs, not one's own. Freedom, at least in its political sense, is coterminous with the absence of bullying or domination. Never­theless, freedom is not the only value that can or should determine behaviour. Moreover to speak of freedom as an end is much too general. I should like to say once again to my critics that the issue is not one between negative freedom as an absolute value and other, inferior, values. It is more complex and more painful. One freedom may abort another; one freedom may obstruct or fail to create conditions which make other freedoms, or a larger degree of freedom, or freedom for other persons, possible; positive and negative freedom may collide; the freedom of the individual or the group may not be fully compatible with a full degree of participation in a common life, with its demands for co-operation, solidarity, fraternity. But beyond all these there is an acuter issue: the paramount need to satisfy the claims of other, no less ultimate, values: justice, happiness, love, the realisation of capacities to create new things and experiences and ideas, the discovery of the truth. Nothing is gained by identifying freedom proper, in either of its senses, with these values, or with the conditions of freedom, or by confounding types of freedom with one another. The fact that given examples of negative freedom (especially where they coincide with powers and rights) - say the freedom of parents or schoolmasters to determine the education of children, of employers to exploit or dismiss their workers, of slave-owners to dispose of their slaves, of the torturer to inflict pain on his victims - may, in many cases, be wholly undesirable, and should in any sane or decent society be curtailed or suppressed, does not render them genuine freedoms any the less; nor does that fact justify us in so reformulating the definition of freedom that it is always repre­sented as something good without qualification - always leading to the best possible consequences, always likely to promote my 'highest' self, always in harmony with the true laws of my own 'real' nature or those of my society, and so on, as has been done in many a classical exposition of freedom, from Stoicism to the social doctrines of our day, at the cost of obscuring profound differences.

If either clarity of thought or rationality in action is not to be hopelessly compromised, such distinctions are of critical import­ance. Individual freedom may or may not clash with democratic organisation, and the positive liberty of self-realisation with the negative liberty of non-interference. Emphasis on negative liberty, as a rule, leaves more paths for individuals or groups to pursue; positive liberty, as a rule, opens fewer paths, but with better reasons or greater resources for moving along them; the two may or may not clash. Some of my critics are made indignant by the thought that a man may, on this view, have more 'negative' liberty under the rule of an easygoing or inefficient despot than in a strenuous, but intolerant, egalitarian democracy. But there is an obvious sense in which Socrates would have had more liberty - at least of speech, and even of action - if, like Aristotle, he had escaped from Athens, instead of accepting the laws, bad as well as good, enacted and applied by his fellow citizens in the democracy of which he possessed, and consciously accepted, full membership. Similarly, a man may leave a vigorous and genuinely 'participatory' democratic State in which the social or political pressures are too suffocating for him, for a climate where there may be less civic participation, but more privacy, a less dynamic and all-embracing communal life, less gregariousness, but also less surveillance. This may appear undesirable to those who look on distaste for public life or social activity as a symptom of malaise, of a deep alienation. But temperaments differ, and too much enthusiasm for common norms can lead to intolerance and disregard for the inner life of man. I understand and share the indignation of democrats; not only because any negative liberty that I may enjoy in an easygoing or inefficient despotism is precarious, or confined to a minority, but because despotism is irrational and unjust and degrading as such: because it denies human rights even if its subjects are not discontented; because participation in self-government, is, like justice, a basic human requirement, an end in itself. Jacobin 'repressive tolerance' destroys individual liberty as effectively as a despotism (however tolerant) destroys positive liberty and degrades its subjects. Those who endure the defects of one system tend to forget the shortcomings of the other. In different historical circumstances some regimes grow more oppressive than others, and to revolt against them is braver and wiser than to acquiesce. Nevertheless, in resisting great present evils, it is as well not to be blinded to the possible danger of the total triumph of any one principle. It seems to me that no sober observer of the twentieth century can avoid qualms in this matter.¹

What is true of the confusion of the two freedoms, or of identifying freedom with its conditions, holds in even greater measure of the stretching of the word 'freedom' to include an amalgam of other desirable things - equality, justice, happiness, knowledge, love, creation and other ends that men seek for their own sakes. This confusion is not merely a theoretical error. Those who are obsessed by the truth that negative freedom is worth little without sufficient conditions for its active exercise, or without the satisfaction of other human aspirations, are liable to minimise its importance, to deny it the very title of freedom, to transfer it to something that they regard as more precious, and finally to forget that without it human life, both social and individual, withers away. If I have been too vehement in the defence of it - only one, I may be reminded, among other human values - and have not insisted as much as my critics demand that to ignore other values can lead to evils at least as great, my insistence upon it in a world in which conditions for freedom may demand an even higher priority does not seem to me to invalidate my general analysis and argument.

Finally one may ask what value there is in liberty as such. Is it a response to a basic need of men, or only something presupposed by other fundamental demands? And further, is this an empirical

' This, indeed, was the point of the penultimate paragraph of Two Concepts of Liberty, which was widely taken as an unqualified defence of 'negative' against 'positive' liberty. This was not my intention. This much criticised passage was meant as a defence, indeed, but of a pluralism based on the perception of incompatibility between the claims of equally ultimate ends, against any ruthless monism which solves such problems by eliminating all but one of the rival claimants. I have therefore revised the text (see pp. 2 16-17 below) to make it clear that I am not offering a blank endorsement of the 'negative' concept as opposed to its 'positive' twin brother, since this would itself constitute precisely the kind of intolerant monism against which the entire argument is directed.

question, to which psychological, anthropological, sociological, historical facts are relevant? Or is it a purely philosophical question, the solution of which lies in the correct analysis of our basic concepts, and for the answer to which the production of examples, whether real or imaginary, and not the factual evidence demanded by empirical enquiries, is sufficient and appropriate? 'Freedom is the essence of man'; 'Frei sein ist nichts - frei werden ist der Himmel' ('To be free is nothing, to become free is very heaven');i 'Every man has a right to life, liberty and the pursuit of happiness.'[33] Do these phrases embody propositions resting on some empirical foundation, or have they some other logical status? Aie they propositions or disguised commands, emotive expres­sions, declarations of intent or commitment? What role, if any, does evidence - historical, psychological, sociological - play in establishing truth or validity in these matters? Could it be the case that if the evidence of the facts should go against us, we should have to revise our ideas, or withdraw them altogether, or at best concede that they - these propositions, if they are propositions - hold only for particular societies, or particular times and places, as some relativists claim?[34] Or is their authority shown by philosoph­ical analysis which convinces us that indifference to freedom is abnormal, that is, offends against what we conceive of as being specifically human, or, at least, fully human - whether by human beings we mean the average members of our own culture, or men in general, everywhere, at all times? To this it is sufficient, perhaps, to say that those who have ever valued liberty for its own sake believed that to be free to choose, and not to be chosen for, is an inalienable ingredient in what makes human beings human; and that this underlies both the positive demand to have a voice in the laws and practices of the society in which one lives, and to be accorded an area, artificially carved out, if need be, in which one is one's own master, a 'negative' area in which a man is not obliged to account for his activities to any man so far as this is compatible with the existence of organised society.

I should like to add one final qualification. Nothing that I assert in the essay on two concepts of liberty about the frontiers of individual liberty (and this applies to the liberty of groups and associations too) should be taken to mean that freedom in any of its meanings is either inviolable, or sufficient, in some absolute sense. It is not inviolable, because abnormal conditions may occur, in which even the sacred frontiers of which Constant speaks, for instance those violated by retrospective laws, punishment of the innocent, judicial murder, information laid against parents by children, the bearing of false witness, may have to be disregarded if some sufficiently terrible alternative is to be averted. Macfarlanei urges this point against me, correctly, it seems to me. Nevertheless, the exception proves the rule: precisely because we regard such situations as being wholly abnormal, and such measures as abhorrent, to be condoned only in emergencies so critical that the choice is between great evils, we recognise that under normal conditions, for the great majority of men, at most times, in most

times have showed little taste for it, and seem contented to be ruled by others, seeking to be well governed by those who provide them with sufficient food, shelter, rules of life, but not to be self-governed. Why should man alone, Herzen asked, be classified in terms of what at most small minorities here or there have ever sought for its own sake, still less actively fought for? This sceptical reflection was uttered by a man whose entire life was dominated by a single-minded passion - the pursuit of liberty, personal and political, of his own and other nations, to which he sacrificed his public career and his private happiness. A. I. Gertsen, Sobranie sochinenii v tridsati tomakh (Moscow, 1954-66), vol. 6, pp. 94^.

¹ op. cit. (p. 32 above, note 1 ).

places, these frontiers are sacred, that is to say, that to overstep them leads to inhumanity. Conversely, the minimum area that men require if such dehumanisation is to be averted, a minimum which other men, or institutions created by them, are liable to invade, is no more than a minimum; its frontiers are not to be extended against sufficiently stringent claims on the part of other values, including those of positive liberty itself. Nevertheless the proper concept of degrees of individual liberty still seems to me to consist in the extent of the area in which choices are open. This minimum area may be incompatible with arrangements required by other social ideals, theocratic or aristocratic or technocratic and the like, but this claim is what the demand for individual liberty entails. Least of all does it call for abdication by individuals or groups from democratic self-government of the society, after their own nicely calculated corner has been made secure and fenced in against others, leaving all the rest to the play of power politics. An indefinite expansion of the area in which men can freely choose between various possible courses of action may plainly not be compatible with the realisation of other values. Hence, things being as they are, we are compelled to adjust claims, compromise, establish priorities, engage in all those practical operations that social and even individual life has, in fact, always required.

If it is maintained that the identification of the value of liberty with the value of a field of free choice amounts to a doctrine of self-realisation, whether for good or evil ends, and that this is closer to positive than to negative liberty, I shall offer no great objection; only repeat that, as a matter of historical fact, distortions of this meaning of positive liberty (or self-determination), even by so well-meaning a liberal as T. H. Green, so original a thinker as Hegel, or so profound a social analyst as Marx, obscured this thesis and at times transformed it into its opposite. Kant, who stated his moral and social position a good deal less equivocally, denounced paternalism, since self-determination is precisely what it obstructs; even if it is indispensable for curing certain evils at certain times, it is, for opponents of tyranny, at best a necessary evil; as are all great accumulations of power as such. Those who maintain¹ that such concentrations are sometimes required to remedy injustices, or to increase the insufficient liberties of individuals or groups, tend to ignore or play down the reverse of the coin: that much power (and

' As do L. J. Macfarlane, ibid., and the majority of democratic theorists.

authority) is also, as a rule, a standing threat to fundamental liberties. All those who have protested against tyranny in modern times, from Montesquieu to the present day, have struggled with this problem. The doctrine that accumulations of power can never be too great, provided that they are rationally controlled and used, ignores the central reason for pursuing liberty in the first place - that all paternalist governments, however benevolent, cautious, disinterested and rational, have tended, in the end, to treat the majority of men as minors, or as being too often incurably foolish or irresponsible; or else as maturing so slowly as not to justify their liberation at any clearly foreseeable date (which, in practice, means at no definite time at all). This is a policy which degrades men, and seems to me to rest on no rational or scientific foundation, but, on the contrary, on a profoundly mistaken view of the deepest human needs.

I have, in the essays that follow, attempted to examine some of the fallacies that rest on misunderstanding of certain central human needs and purposes - central, that is, to our normal notion of what it is to be a human being; a being endowed with a nucleus of needs and goals, a nucleus common to all men, which may have a shifting pattern, but one whose limits are determined by the basic need to communicate with other similar beings. The notion of such a nucleus and such limits enters into our conception of the central attributes and functions in terms of which we think of men and societies.

I am only too fully conscious of some of the difficulties and obscurities which my thesis still contains. But short of writing another book, I could do no more than deal with those criticisms which seemed to me at once the most frequent and the least effective, resting as they do on an over-simple application of particular scientific or philosophical principles to social and political problems. But even here I am well aware of how much more needs to be done, especially on the issue of free will, the solution of which seems to me to require a set of new conceptual tools, a break with traditional terminology, which no one, so far as I know, has yet been able to provide.