tof eofj "ccLAW, LEGISLATION AND LIBERTYThis is Hayek's major statement of political philosophy. RejectingMarx, Freud, logical positivism and political egalitarianism, Hayekshows that the naive application of scientific methods to cultureand education has been harmful and misleading, creating super-stition and error rather than an age of reason and culture.Law, Legislation and Liberty combines all three volumes ofHayek's comprehensive study on the basic principles of thepolitical order of a free society. Rules and Order deals with thebasic conceptions necessary for a critical analysis of prevailingtheories of justice and of conditions which a constitution securingpersonal liberty would have to satisfy. The Mirage of Social Justicepresents a critical analysis of the theories of utilitarianism, legalpositivism and 'social justice'. The Political Order of a Free Peopledemonstrates that the democratic ideal is in danger of miscarryingdue to confusions of egalitarianism and democracy, erroneousassumptions that there can be moral standards without moral disci-pline, and that tradition can be ignored in proposals for restruc-turing society.F.A. Hayek became both a Doctor of Law and a Doctor of Poli-tical Science at the University of Vienna. He was made the firstDirector of the Austrian Institute of Economic Research and in1931 was appointed to a chair at the London School of Econ-omics. In 1950 he went to the University of Chicago as Professorof Social and Moral Sciences and then became Professor of Econ-omics at the Albert-Ludwigs-Universitat of Frieburg and ProfessorEmeritus in 1967. He was also a Fellow of the British Academyand was awarded a Nobel Memorial Prize in Economics in 1974.Hayek died in 1992.LAW, LEGISLATIONAND LIBERTYA new statement of the liberal principlesof justice and political economyVolume 1RULES AND ORDERVolume 2THE MIRAGE OF SOCIAL JUSTICEVolume 3THE POLITICAL ORDEROF A FREE PEOPLEF. A. HayekVol. 1 Rules and Order first published 1973Vol. 2 The Mirage of Social Justice first published 1976Vol. 3 The Political Order of a Free People first published 1979First published in one volume with corrections and revised prefacein 1982 by Routledge & Kegan Paul Ltd.Reprinted 1993, 1998by Routledge11 New Fetter Lane, London EC4P 4EE F. A. Hayek 1973, 1976, 1979, 1982Printed and bound in Great Britain byT.l. International Ltd, Padstow, CornwallAll rights reserved. No part of this book may be reprinted orreproduced or utilised in any form or by any electronic,mechanical, or other means, now known or hereafter invented,including photocopying and recording, or in any informationstorage or retrieval system, without permission in writing from thepublishers.British Library Cataloguing in Publication DataA catalogue record for this book is available from the BritishLibraryISBN 0-415-09868-8CONTENTSVolume 1RULES AND ORDERCONSOLIDATED PREFACEINTRODUCTIONxvREASON AND EVOLUTION 8Construction and evolution 8The tenets ofCartesian rationalism 9The permanent limitations of ourfactual knowledge 11Factual knowledge and science 15The concurrent evolution of mind and society: the roleof rules 17The false dichotomy of 'natural' and 'artificial' 20The rise of the evolutionary approach 22The persistence ofconstructivism in current thought 24Our anthropomorphic language 26Reason and abstraction 29Why the extreme forms ofconstructivist rationalismregularly lead to a revolt against reason 312 COSMOS AND TAXIS 35The concept of order 35The two sources of order 36The distinguishing properties ofspontaneous orders 38Spontaneous orders in nature 39In society, reliance on spontaneous order both extends andlimits our powers ofcontrol 41Spontaneous orders result from their elements obeyingcertain rules ofconduct 43The spontaneous order ofsociety is made up of individualsand organizations 46vCONTENTSThe rules ofspontaneous orders and the rules oforganization 48The terms 'organism' and 'organization' 523 PRINCIPLES AND EXPEDIENCY 55Individual aims and collective benefits 55Freedom can be preserved only by following principles andis destroyed by following expediency 56The 'necessities' ofpolicy are generally the consequencesofearlier measures 59The danger ofattaching greater importance to the predictablerather than to the merelypossibleconsequences ofouractions 61Spurious realisln and the requiredcourage to consider utopia 62The role of the lawyer in political evolution 65The modern development of law has been guided largely byfalse economics 674 THE CHANGING CONCEPT OF LAW 72Law is older than legislation 72The lessons ofethology and cultural anthropology 74The process 0.[ articulation ofpractices 76Factual and normative rules 78Early law 81The classical and the medieval tradition 82The distinctive attributes oflaw arising from custom andprecedent 85Why grown law requires correction by legislation 88The origin of legislative bodies 89Allegiance and sovereignty 915 NOMOS: THE LAW OF LIBERTY 94The functions of the judge 94How the task of the judge differs froIn that of the head ofan organization 97The aiJn ofjurisdiction is the Inaintenance of an ongoingorder of actions 98'Actions towards others' and theprotection ofexpectations 101viCONTENTSIn a dynamic order ofactions only some expectations canbe protected 102The maximal coincidence ofexpectations is achieved bythe deli/nitation ofprotected domains 106The general problem of the effects of values on facts 110The 'purpose' of law 112The articulations of the law and the predictability ofjudicial decisions 115Thefunction ofthejudge is confinedto aspontaneous order 118Conclusions 1226 THESIS: THE LAW OF LEGISLATION 124Legislation originates from the necessity of establishingrules of organization 124Law and statute-the enforcement of law and the executionofcommands 126Legislation and the theory of the separation ofpowers 128The governmental functions of representative asselnblies 129Private law and public law 131Constitutional law 134Financial legislation 136Administrative law and the police power 137The 'Ineasures , ofpolicy 139The transformation ofprivate law into public law by'social'legislation 141The Inental bias ofa legislaturepreoccupied with governlnent 143NOTESvii145CONTENTSVolume 2THE MIRAGE OF SOCIAL JUSTICE7 GENERAL WELFARE AND PARTICULAR PURPOSESIn afree society the general good consists principally inthe facilities for the pursuit of unknown purposes 1The general interest and collective goods 6Rules and ignorance 8The significance ofabstract rules in a world in which mostof the particulars are unknown 11Will and opinion, ends and values, commands and rules,and other terminological issues 12Abstract rules operate as ultimate values because theyserve unknown particular ends 15The constructivist fallacy of utilitarianism 17All valid criticism or improvement of rules of conductmust proceed within a given system ofrules 24'Generalization' and the test of universalizabiiity 27To perform their functions rules must be appliedthroughout the long run 298 THE QUEST FOR JUSTICE 31Justice is an attribute of human conduct 31Justice and the law 34Rules ofjust conduct are generally prohibitions of unjustconduct 35Not only the rules ofjust conduct, but also the test oftheir justice, are negative 38The significance of the negative character of the test ofinjustice 42The ideology of legal positivism 44The 'pure theory oflaw' 48viiiCONTENTSLaw and morals 56The 'law ofnature' 59Law and sovereignty 619 'SOCIAL' OR DISTRIBUTIVE JUSTICE 62The concept of 'social justice' 62The conquest ofpublic imagination by 'social justice' 65The inapplicability of the concept ofjustice to theresults ofa spontaneous process 67The rationale of the economic game in which only theconduct of the players but not the result can be just 70The alleged necessity ofa belief in the justice of rewards 73There is no 'value to society' 75The meaning of 'social' 78'Social justice' and equality 80'Equality of opportunity' 84'Social justice' andfreedom under the law 85The spatial range of 'social justice' 88Claims for compensation for distasteful jobs 91The resentment of the loss of accustomed positions 93Conclusions 96APPENDIX TO CHAPTER 9 JUSTICE ANDINDIVIDUAL RIGHTS 101lOTHEMARK ETORDER 0 RCATAL L A X Y 107The nature of the market order 107A free society is a pluralistic society without a commonhierarchy ofends 109Though not asingle economy, the Great Society is still heldtogether by what vulgarly are calledeconomic relations 112The aim ofpolicy in a society offree men cannot be amaximumofforeknown results but only an abstract order 114The game ofcatallaxy 115In judging the adaptations to changing circumstancescomparisons of the new with the former position areirrelevant 120ixCONTENTSRules ofjust conduct protect only material domains andnot market values 123The correspondence ofexpectations is brought about by adisappointment ofsome expectations 124Abstract rules ofconduct can determine only chances andnot particular results 126Specific comlnands ('interference') in a catallaxy createdisorder and can never be just 128The aim oflaw should be to improve equally the chancesofall 129The Good Society is one in which the chances of anyoneselected at random are likely to be as great as possible 13211 THE DISCIPLINE OF ABSTRACT RULES AND THEEMOTIONS OF THE TRIBAL SOCIETY 133The pursuit of unattainable goals may prevent theachievement of the possible 133The causes of the revival of the organizational thinkingof the tribe 134The immoral consequences of morally inspired efforts 135In the Great Society 'social justice' becomes a disruptiveforce 137From the care of the most unfortunate to the protectionof vested interests 139Attempts to 'correct' the order of the market lead to itsdestruction 142The revolt against the discipline ofabstract rules 143The morals of the open and of the closed society 144The old conflict between loyalty andjustice 147The small group in the Open Society 149The importance of voluntary associations 150NOTES 153xCONTENTSVolume 3THE POLITICAL ORDER OF AFREE PEOPLE12 MAJORITY OPINION AND CONTEMPORARYDEMOCRACYThe progressive disillusionment about democracyUnlimited power the fatal effect of the prevailing formofdemocracy 3The true content of the democratic ideal 5The weakness ofan elective assembly with unlimitedp o w e ~ 8Coalitions of organized interests and the apparatus ofpara-government 13Agreement on general rules and on particular measures 1713 THE DIVISION OF DEMOCRATIC POWERS 20The loss of the original conception of the functions ofalegislature 20Existing representative institutions have been shaped bythe needs ofgovernment, not of legislation 22Bodies with powers ofspecific direction are unsuitedforlaw-making 25The character ofexisting 'legislatures' determined by theirgovernmental tasks 27Party legislation leads to the decay ofdemocratic society 31The constructivistic superstition ofsovereignty 33The requisite division of the powers ofrepresentativeassemblies 35Democracy or demarchy? 38xiCONTENTS14 THE PUBLIC SECTOR AND THE PRIVATESECTOR 41The double task ofgovernment 41Collective goods 43The delimitation of the public sector 46The independent sector 49Taxation and the size of the public sector 51Security 54Government monopoly ofservices 56Information and education 60Other critical issues 6215 GOVERNMENT POLICY AND THE MARKET 65The advantages ofcompetition do not depend on it being'perfect' 65Competition as a discovery procedure 67If the factual requirements of 'perfect' competition areabsent, it is not possible to makefirms act 'as if' it existed 70The achievements of the free market 74Competition and rationality 75Size, concentration and power 77The political aspects ofeconomic power 80When monopoly becomes harmful 83The problem of anti-monopoly legislation 85Not individual, but group selfishness is the chief threat 89The consequences ofa political determination of theincomes of the different groups 93Organizable and non-organizable interests 9616 THE MISCARRIAGE OF THE DEMOCRATICIDEAL: A RECAPITUALATION 98The miscarriage of the democratic ideal 98A 'bargaining' democracy 99The playball ofgroup interests 99Laws versus directions 100Laws and arbitrary government 101Froln unequal treatment to arbitrariness 102Separation ofpowers to prevent unlimited governlnent 104xiiCONTENTS17 A MODEL CONSTITUTION 105The wrong turn taken by the development ofrepresentativeinstitutions 105The value ofa model of an ideal constitution 107The basic principles 109The two representative bodies with distinctive functions 111Further observations on representation by age groups 117The governmental assembly 119The constitutional court 120The general structure ofauthority 122Emergency powers 124The division offinancial powers 12618 THE CONTAINMENT OF POWER AND THEDETH RONEM ENT OF POL ITICS 128Lilnited and unlimited power 128Peace, freedom andjustice: the three great negatives 130Centralization and decentralization 132The rule of the Inajority versus the rule of laws approvedby the majority 133Moral confusion and the decay of language 135Democratic procedure and egalitarian objectives 137'State' and 'society' 139A game according to rules can never knowjustice oftreatment 141The para-government of organized interests and thehypertrophy ofgovernment 143Unlimited democracy and centralization 145The devolution of internal policy to local government 146The abolition of the government monopoly ofservices 147The dethronement ofpolitics 149EPILOGUE: THE THREE SOURCES OF HUMANVALUES 153The errors ofsociobiology 153The process of cultural evolution 155The evolution ofself-maintaining complex systems 158The stratification ofrules ofconduct 159xiiiCONTENTSCustomary rules and economic order 161The discipline offreedom 163The re-emergence ofsuppressed primordial instincts 165Evolution, tradition and progress 168The construction of new morals to serve old instincts:A1arx 169The destruction ofindispensable values byscientific error:Freud 173The tables turned 175NOTES 177I NDE X 0 F AUT H 0 RSCI TED I N VOL U M ES 1- 3 209SUBJECT INDEX TO VOLUMES 1-3 217xivCONSOLIDATED PREFACETO ONE-VOLUME EDITIONAt last this work can appear in the form it was intended to takewhen I started on it nearly twenty years ago. Half way through thisperiod, when a first draft was nearly completed, a weakening of mypowers, which fortunately proved to be temporary, made me doubtwhether I should ever be able to complete it and led me to publishin 1973 a fully completed part of what were to become threeseparate volumes. When a year later I found my powers returning Idiscovered that various circumstances made substantial revisionsnecessary of even those further parts of the draft which I hadthought to be in fairly finished state. As I explained in the prefaceto the second volume, which appeared in 1976, the chief reason wasmy dissatisfaction with that central chapter which gave that volumeits sub-title The Mirage of Social Justice. This account] had betterrepeat here:I had devoted to this subject an enormous chapter in which Ihad tried to show for a large number of instances that whatwas claimed as demanded by 'social justice' could not bejustice because the underlying consideration (one could hardlycall it a principle) was not capable of general application. Thepoint I was then mainly anxious to demonstrate was thatpeople would never be able to agree on what 'social justice'required, and that any attempt to determine remunerationsaccording to what it was thought was demanded by justicewould make the market unworkable. I have now becomeconvinced, however, that the people who habitually employthe phrase simply do not know themselves what they mean byit and just use it as an assertion that a claim is justified'without giving a reason for it.In my earlier efforts to criticize the concept I had all thetime the feeling that I was hitting into a void and I finallyattempted, what in such cases one ought to do in the firstxvPREFACEinstance, to construct as good a case in support of the ideal of'social justice' as was in my power. It was only then that Iperceived that the Emperor had no clothes on, that is, that theterm 'social justice' was entirely empty and meaningless. Asthe boy in Hans Christian Andersen's story, I 'could not seeanything, because there was nothing to be seen.' The more Itried to give it a definite meaning the more it fell apart-theintuitive feeling of indignation which we undeniably oftenexperience in particular instances proved incapable of beingjustified by a general rule such as the conception of justicedemands. But to demonstrate that a universally usedexpression which to many people embodies a quasi-religiousbelief has no content whatever and serves merely to insinuatethat we ought to consent to a demand of some particulargroup is much more difficult than to show that a conceptionis wrong.In these circumstances I could not content myself to showthat particular attempts to achieve 'social justice' would notwork, but had to explain that the phrase meant nothing at all,and that to employ it was either thoughtless or fraudulent. Itis not pleasant to have to argue against a superstition which isheld most strongly by men and women who are oftenregarded as the best in our society, and against a belief thathas become almost the new religion of our time (and in whichmany of the ministers of old religion have found their refuge),and which has become the recognized mark of the good man.But the present universality of that belief proves no more thereality of its object than did the universal belief in witches orthe philosopher's stone. Nor does the long history of theconception of distributive justice understood as an attribute ofindividual conduct (and now often treated as synonymouswith 'social justice') prove that it has any relevance to thepositions arising from the market process. I believe indeedthat the greatest service I can still render to my fellow menwould be if it were in my power to make them ashamed ofever again using that hollow incantation. I felt it my duty atleast to try and free them of that incubus which today makesfine sentiments the instruments for the destruction of allvalues of a free civilization-and to try this at the risk ofgravely offending many the strength of whose moral feelings Irespect.xviPREFACEThe present version of the central chapter of this volumehas in consequence of this history in some respects a slightlydifferent character from the rest of the volume which in allessentials was completed six or seven years earlier. There was,on the one hand, nothing I could positively demonstrate butmy task was to put the burden of proof squarely on thosewho employ the term. On the other hand, in re-writing thatchapter I no longer had that easy access to adequate libraryfacilities which I had when I prepared the first draft of thisvolume. I have in consequence not been able in that chaptersystematically to take account of the more recent literature onthe topics I discussed as I had endeavoured to do in the restof this volume. In one instance the feeling that I ought tojustify my position vis-a-vis a major recent work has alsocontributed to delay the completion of this volume. But aftercareful consideration I have come to the conclusion that whatI might have to say about John Rawls' A Theory of Justice(1972) would not assist in the pursuit of my immediate objectbecause the differences between us seemed more verbal thansubstantial. Though the first impression of readers may bedifferent, Rawls' statement which I quote later in this volume(p. 100) seems to me to show that we agree on what is to methe essential point. Indeed, as I indicate in a note to thatpassage, it appears to me that Rawls has been widelymisunderstood on this central issue.The preface to the third volume, which ultimately appeared in1979, gives a similar account of the further development that alsohad better be repeated here:Except for what are now the last two chapters, most of it wasin fairly finished form as long ago as the end of 1969 whenindifferent health forced me to suspend the efforts tocomplete it. It was then, indeed, doubt whether I would eversucceed in doing so which made me decide to publishseparately as volume 1 the first third of what had beenintended to form a single volume, because it was incompletely finished form. When I was able to return tosystematic work I discovered, as I have explained in thepreface to volume 2, that at least one chapter of the originaldraft of that part required complete re-writing.Of the last third of the original draft only what wasxviiPREFACEintended to be the last chapter (chapter 18) had not beencompleted at the time when I had discontinued work. Butwhile I believe I have now more or less carried out theoriginal intention, over the long period which has elapsed myideas have developed further and I was reluctant to send outwhat inevitably must be my last systematic work without atleast indicating in what direction my ideas have been moving.This has had the effect that not only what was meant to bethe concluding chapter contains a good deal of, I hope,improved re-statements of arguments I have developed earlier,but that I found it necessary to add an Epilogue whichexpresses more directly the general view of moral and politicalevolution which has guided me in the whole enterprise. I havealso inserted as chapter 16 a brief recapitulation of the earlierargument.There were also other causes which have contributed todelay completion. As I had hesitated whether I ought topublish volume 2 without taking full account of the importantwork of John Rawls, A Theory of Justice (Oxford, 1972), twonew important books in the field have since appeared which,if I were younger, I should feel I must fully digest beforecompleting my own survey of the same kind of problems:Robert Nozik, Anarchy, State and Utopia (New York, 1974)and Michael Oakeshott, On Human Conduct (Oxford, 1975).Rightly or wrongly I finally decided that if I made an effortfully to absorb their argument before concluding my ownexposition, I would probably never do this. But I regard it asmy duty to tell the younger readers that they cannot fullycomprehend the present state of thought on these issues unlessthey make that effort which I must postpone until I havecompleted the statement of the conclusions at which I hadarrived before I became acquainted with these works.The long period over which the present work has beengrowing also had the effect that I came to regard it asexpedient to change my terminology on some points on whichI should warn the reader. It was largely the growth ofcybernetics and the related subjects of information and systemtheory which persuaded me that expression other than thosewhich I habitually used may be more readily comprehensibleto the contemporary reader. Though I still like andoccasionally use the term 'spontaneous order', I agree thatxviiiPREFACE'self-generating order' or 'self-organizing structures' aresometimes more precise and unambiguous and thereforefrequently use them instead of the former term. Similarly,instead of 'order', in conformity with today's predominantusage, I occasionally now use 'system'. Also 'information' isclearly often preferable to where I usually spoke of'knowledge', since the former clearly refers to the knowledgeof particular facts rather than theoretical knowledge to whichplain 'knowledge' might be thought to refer. Finally, since'constructivist' appears to some people still to carry thecommendatory connotation derived from the adjective'constructive', I felt it advisable, in order clearly to bring outthe deprecatory sense in which I use that term (significantly ofRussian origin) to employ instead the, I am afraid, still moreugly term 'constructivistic'. I should perhaps add that I feelsome regret that I have not had the courage consistently toemploy certain other neologisms I had suggested, such as'cosmos', 'taxis', 'nomos', 'thesis', 'catallaxy' and'demarchy'. But what the exposition has thereby lost inprecision it will probably have gained in ready intelligibility.Perhaps I should also again remind the reader that thepresent work was never intended to give an exhaustive orcomprehensive exposition of the basic principles on which asociety of free man could be maintained, but was rathermeant to fill the gaps which I discovered after I had made anattempt to restate, in The Constitution of Liberty, for thecontemporary reader the traditional doctrines of classicalliberalism in a form suited to contemporary problems andthinking. It is for this reason a much less complete, muchmore difficult and personal but, I hope, also more originalwork than the former. But it is definitely supplementary toand not a substitute for it. To the non-specialist reader Iwould therefore recommend reading The Constitution ofLiberty before he proceeds to the more detailed discussion orparticular examination of problems to which I have attemptedsolutions in these volumes. But they are intended to explainwhy I still regard what have now long been treated asantiquated beliefs as greatly superior to any alternativedoctrines which have recently found more favour with thepublic.The reader will probably gather that the whole work hasxixPREFACEbeen inspired by a growing apprehension about the directionin which the political order of what used to be regarded as themost advanced countries is teuding. The growing conviction,for which the book gives the reasons, that this threateningdevelopment towards a totalitarian state is made inevitable bycertain deeply entrenched defects of construction of thegenerally accepted type of 'democratic' government hasforced me to think through alternative arrangements. Iwould like to repeat here that, though I profoundly believein the basic principles of democracy as the only effectivemethod which we have yet discovered of making peacefulchange possible, and am therefore much alarmed by theevident growing disillusionment about it as a desirable Inelhodof government-much assisted by the increasing abuseof the word to indicate supposed ailns of government-I am becoming more and more convinced that weare moving towards an impasse from which politicalleaders will offer to extricate us by desperate means.When the present volume leads up to a proposal of basicalteration of the structure of democratic government, which atthis time most people will regard as wholly impractical, this ismeant to provide a sort of intellectual stand-by equipment forthe time, which may not be far away, when the breakdown ofthe existing institutions becomes unmistakable and when Ihope it may show a way out. It should enable us to preservewhat is truly valuable in democracy and at the same time freeus of its objectionable features which most people still acceptonly because they regard them as inevitable. Together with thesimilar stand-by scheme I have proposed for deprivinggovernment of the monopolistic powers of control of thesupply of money, equally necessary if we are to escape thenightmare of increasingly totalitarian powers, which I haverecently outlined in another publication (Denationalisation ofMoney, 2nd edn, Institute of Economic Affairs, London,1978), it proposes what is a possible escape from the fatewhich threatens us. I shall be content if I have persuadedsome people that if the first experiment of freedom we havetried in modern times should prove a failure, it is not becausefreedom is an impracticable ideal, but because we have tried itthe wrong way.xxPREFACEI trust the reader will forgive a certain lack of system andsome unnecessary repetitions in an exposition which has beenwritten and re-written over a period of fifteen years, brokenby a long period of indifferent health. I am very much awareof this, but if I tried in my eightieth year to recast it all, Ishall probably never complete the task.The Epilogue I added to that volume before publication indicatesthat even during the period of restricted activity my ideas havecontinued to develop imperceptibly more than I was aware before Iattempted to sketch my present general view of the whole positionin a public lecture. As I said in the concluding words of the presenttext, it became clear to me that what I said in that Epilogue shouldnot be an Epilogue but a new beginning. I am glad to be able to saynow that it has turned out to be such and that that Epilogue hasbecome the outline of a new book of which I have now completed afirst draft.There are a few acknowledgments that I ought to repeat here. Someten years ago Professor Edwin McClellan of the University ofChicago had again, as on earlier occasions, taken great trouble tomake my exposition more readable than I myself could have done.I am deeply grateful for his sympathetic efforts but should add,that since even in the early parts the draft on which he has workedhas since undergone further change, he must not be heldresponsible for whatever defects the present version still has. I havehowever incurred further obligations to Professor Arthur Shenfieldof London who has gone through the final text of the third volumeand corrected there a variety of substantial as well as stylisticpoints, and to Mrs Charlotte Cubitt who, in preparing the finalcopy of that volume, has further polished the text. I am also muchindebted to Mrs Cornelia Crawford of Irvington-on-Hudson, NewYork, who has again applied her proven skill and understanding inpreparing the subject index giving references to all three stillseparately paginated volumes.xxiLAW, LEGISLATION AND LIBERTYVolume 1RULES AND ORDERIntelligent beings may have laws of their own making; but theyalso have some which they never made.(Montesquieu, De l'Esprit des lois, I, p. i)INTRODUCTIONThere seems to be only one solution to the problem: that theelite of mankind acquire a consciousness of the limitation ofthe human mind, at once simple and profound enough, humbleand sublime enough, so that Western civilisation will resignitself to its inevitable disadvantages.G. Ferrero*When Montesquieu and the framers of the American Constitutionarticulated the conception of a limiting constitution1 that hadgrown up in England, they set a pattern which liberal constitu-tionalism has followed ever since. Their chief aim was to provideinstitutional safeguards of individual freedom; and the device inwhich they placed their faith was the separation of powers. In theform in which we know this division of power between the legisla-ture, the judiciary, and the administration, it has not achievedwhat it was meant to achieve. Governments everywhere have ob-tained by constitutional means powers which those men had meantto deny them. The first attempt to secure individual liberty byconstitutions has evidently failed.Constitutionalism means limited government. 2 But the interpre-tation given to the traditional formulae of constitutionalism hasmade it possible to reconcile these with a conception of democracyaccording to which this is a form of government where the will ofthe majority on any particular matter is unlimited. 3As a result ithas already been seriously suggested that constitutions are an anti-quated survival which have no place in the modern conception ofgovernment. 4And, indeed, what function is served by a constitu-tion which makes omnipotent government possible? Is its functionto be merely that governments work smoothly and efficiently,whatever their aims?In these circumstances it seems important to ask what thosefounders of liberal constitutionalism would do today if, pursuingINTRODUCTIONthe aims they did, they could command all the experience we havegained in the meantime. There is much we ought to have learnedfrom the history of the last two hundred years that those men withall their wisdom could not have known. To me their aims seem tobe as valid as ever. But as their means have proved inadequate,new institutional invention is needed.In another book I have attempted to restate, and hope to have insome measure succeeded in clarifying, the traditional doctrine ofliberal constitutionalism.5But it was only after I had completedthat work that I came to see clearly why those ideals had failed toretain the support of the idealists to whom all the great politicalmovements are due, and to understand what are the governing be-liefs of our time which have proved irreconcilable with them. Itseems to me now that the reasons for this development were chiefly:the loss of the belief in a justice independent of personal interest; aconsequent use of legislation to authorize coercion, not merely toprevent unjust action but to achieve particular results for specificpersons or groups; and the fusion in the same representative assem-blies of the task of articulating the rules of just conduct with that ofdirecting government.What led me to write another book on the same general theme asthe earlier one was the recognition that the preservation of asociety of free men depends on three fundamental insights whichhave never been adequately expounded and to which the three mainparts of this book are devoted. The first of these is that a self-generating or spontaneous order and an organization are distinct,and that their distinctiveness is related to the two different kinds ofrules or laws which prevail in them. The second is that what todayis generally regarded as 'social' or distributive justice has meaningonly within the second of these kinds of order, the organization;but that it is meaningless in, and wholly incompatible with, thatspontaneous order which Adam Smith called 'the Great Society',and Sir Karl Popper called 'the Open Society'. The third is that thepredominant model of liberal democratic institutions, in which thesan1e representative body lays down the rules of just conduct anddirects government, necessarily leads to a gradual transformation ofthe spontaneous order of a free society into a totalitarian systemconducted in the service of some coalition of organized interests.This development, as I hope to show, is not a necessary conse-quence of democracy, but an effect only of that particular form ofunlimited government vvith which delllocracy has come to be identi-2INTRODUCTIONfied. If I aln right, it would indeed seem that the particular form ofrepresentative government which now prevails in the Westernworld, and \vhich many feel they must defend because they nlis-takenly regard it as the only possible form of democracy, has an in-herent tendency to lead away from the ideals it was intended toserve. It can hardly be denied that, since this type of democracyhas come to be accepted, we have been moving away from that idealof individual liberty of which it had been regarded as the surestsafeguard, and are now drifting towards a system " , ~ h i c h nobodywanted.Signs are not wanting, however, that unlimited democracy isriding for a fall and that it will go down, not with a bang, but witha whimper. It is already becoming clear that many of the expecta-tions that have been raised can be met only by taking the powers ofdecision out of the hands of democratic assemblies and entrustingthem to the established coalitions of organized interests and theirhired experts. Indeed, we are already told that the function ofrepresentative bodies has become to 'mobilize consent', 6 that is,not to express but to manipulate the opinion of those whom theyrepresent. Sooner or later the people will discover that not only arethey at the mercy of new vested interests, but that the politicalmachinery of para-government, which has grown up as a necessaryconsequence of the provision-state, is producing an impasse bypreventing society from making those adaptations which in achanging world are required to maintain an existing standard ofliving, let alone to achieve a rising one. It will probably be sometime before people will admit that the institutions they have createdhave led them into such an impasse. But it is probably not tooearly to begin thinking about a way out. And the conviction that thiswill demand some drastic revision of beliefs now generally accep-ted is what makes me venture here on some institutional invention.If I had known when I published The Constitution of Libertythat I should proceed to the task attempted in the present work, Ishould have reserved that title for it. I then used the term 'consti-tution' in the wide sense in which we use it also to describe thestate of fitness of a person. It is only in the present book that Iaddress myself to the question of what constitutional arrange-ments, in the legal sense, might be most conducive to the preserva-tion of individual freedom. Except for a bare hint which fe\v readerswill have noticed,7 I confined myself in the earlier book to statingthe principles which the existing types of government would have3INTRODUCTIONto follow if they wished to preserve freedom. Increasing awarenessthat the prevailing institutions make this impossible has led me toconcentrate more and more on what at first seemed merely anattractive but impracticable idea, until the utopia lost its strange-ness and came to appear to me as the only solution of the problemin which the founders of liberal constitutionalism failed.Yet to this problem of constitutional design I turn only in volume3 of this work. To make a suggestion for a radical departure fromestablished tradition at all plausible required a critical re-examina-tion not only of current beliefs but of the real meaning of somefundamental conceptions to which we still pay lip-service. In fact,I soon discovered that to carry out what I had undertaken wouldrequire little less than doing for the twentieth century what Montes-quieu had done for the eighteenth. The reader will believe me whenI say that in the course of the work I more than once despaired ofmy ability to come even near the aim I had set myself. I am notspeaking here of the fact that Montesquieu was also a great literarygenius whom no mere scholar can hope to emulate. I refer ratherto the purely intellectual difficulty which is a result of the circum-stance that, while for Montesquieu the field which such an under-taking must cover had not yet split into numerous specialisms, ithas since become impossible for any man to master even the mostimportant relevant works. Yet, although the problem of an appro-priate social order is today studied from the different angles ofeconomics, jurisprudence, political science, sociology, and ethics,the problem is one which can be approached successfully only as awhole. This means that whoever undertakes such a task today can-not claim professional competence in all the fields with which hehas to deal, or be acquainted with the specialized literature avail-able on all the questions that arise.Nowhere is the baneful effect of the division into specialismsmore evident than in the two oldest of these disciplines, economicsand law. Those eighteenth-century thinkers to whom we owe thebasic conceptions of liberal constitutionalism, David Hume andAdam Smith, no less than Montesquieu, were still concerned withwhat some of them called the 'science of legislation', or with princi-ples of policy in the widest sense of this term. One of the mainthemes of this book will be that the rules of just conduct which thelawyer studies serve a kind of order of the character of which thelawyer is largely ignorant; and that this order is studied chiefly bythe economist who in turn is similarly ignorant of the character of4INTRODUCTIONthe rules of conduct on which the order that he studies rests.The most serious effect of the splitting up among several specia-lisms of what was once a common field of inquiry, however, is thatit has left a no-man's-land, a vague subject sometimes called'social philosophy'. Some of the chief disputes within those specialdisciplines turn, in fact, on differences about questions which arenot peculiar to, and are therefore also not systematically examinedby, anyone of them, and which are for this reason regarded as'philosophical'. This serves often as an excuse for taking tacitly aposition which is supposed either not to require or not to be capableof rational justification. Yet these crucial issues on which not onlyfactual interpretations but also political positions wholly depend,are questions which can and must be answered on the basis of factand logic. They are 'philosophical' only in the sense that certainwidely but erroneously held beliefs are due to the influence of aphilosophical tradition which postulates a false answer to questionscapable of a definite scientific treatment.In the first chapter of this book I attempt to show that certainwidely held scientific as well as political views are dependent on aparticular conception of the formation of social institutions, whichI shall call 'constructivist rationalism'-aconception which assumesthat all social institutions are, and ought to be, the product ofdeliberate design. This intellectual tradition can be shown to befalse both in its factual and in its normative conclusions, becausethe existing institutions are not all the product of design, neitherwould it be possible to make the social order vvholly dependent ondesign without at the same time greatly restricting the utilization ofavailable knowledge. That erroneous view is closely connectedwith the equally false conception of the human mind as an entitystanding outside the cosmos of nature and society, rather thanbeing itself the product of the same process of evolution to whichthe institutions of society are due.I have indeed been led to the conviction that not only some ofthe scientific but also the most important political (or 'ideological')differences of our time rest ultimately on certain basic philosophi-cal differences between two schools of thought, of which one canbe shown to be mistaken. They are both commonly referred to asrationalism, but I shall have to distinguish between them as theevolutionary (or, as Sir Karl Popper calls it, 'critical') rationalismon the one hand, and the erroneous constructivist (Popper's'naIve') rationalism on the other. If the constructivist rationalism5INTRODUCTIONcan be sho\vn to be based on factually false assumptions, a wholefamily of schools of scientific as well as political thought will also beproved erroneous.In the theoretical fields it is particularly legal positivisn1 and theconnected belief in the necessity of an unlimited 'sovereign' po\verwhich stand or fall \vith this error. The same is true of utilitari-anism, at least in its particularistic or 'act' variety; also, I am afraidthat a not inconsiderable part of what is called 'sociology' is adirect child of constructivisn1 when it presents its aims as 'to createthe future of mankind' 8 or, as one writer put it, claims 'that socialismis the logical and inevitable outcome of sociology'. 9 All the totali-tarian doctrines, of \vhich socialism is merely the noblest and mostinfluential, indeed belong here. They are false, not because of thevalues on \vhich they are based, but because of a misconception ofthe forces \vhich have Inade the Great Society and civilizationpossible. r-rhe demonstration that the differences between socialistsand non-socialists ultimately rest on purely intellectual issuescapable of a scientific resolution and not on different judgments ofvalue appears to me one of the most important outcomes of thetrain of thought pursued in this book.It appears to me also that the same factual error has long appearedto make insoluble the most crucial problem of political organiza-tion, namely ho\" to limit the 'popular will' \vithout placing another'"rill' above it. As soon as \ve recognize that the basic order of theGreat Society cannot rest entirely on design, and can therefore alsonot aim at particular foreseeable results, we see that the require-ment, as legitilnation of all authority, of a commitment to generalprinciples approved by general opinion, Inay well place effectiverestrictions on the particular \yill of all authority, including that ofthe Inajority of the rnoment.On these issues \vhich \vill be my main concern, thought seemsto have made little advance since David Hume and Imlnanuel Kant,and in several respects it \vill be at the point at which they left offthat our analysis will have to resume. It was they who came nearerthan anybody has done since to a clear recognition of the status ofvalues as independent and guiding conditions of all rational con-struction. What I am ultimately concerned with here, although I candeal only \vith a small aspect of it, is that destruction of values byscientific error which has increasingly come to seem to me the greattragedy of our time-a tragedy, because the values which scientificerror tends to dethrone are the indispensable foundation of all our6INTRODUCTIONcivilization, including the very scientific efforts which have turnedagainst them. The tendency of constructivism to represent thosevalues which it cannot explain as determined by arbitrary human- decisions, or acts of will, or mere emotions, rather than as the neces-sary conditions of facts which are taken for granted by its expoun-ders, has done much to shake the foundations of civilization, and ofscience itself, which also rests on a system of values which cannotbe scientifically proved.7ONEREASON AND EVOLUTIONTo relate by whom, and in what connection, the true law ofthe formation of free states was recognized, and how thisdiscovery, closely akin to those which, under the names ofdevelopment, evolution, and continuity, have given a new anddeeper method to other sciences, solved the ancient problembet\veen stability and change, and determined the authority oftradition on the progress of thought.Lord Acton*Construction and evolutionThere are two ways of looking at the pattern of human activitieswhich lead to very different conclusions concerning both its expla-nation and the possibilities of deliberately altering it. Of these, oneis based on conceptions which are demonstrably false, yet are sopleasing to human vanity that they have gained great influence andare constantly employed even by people who know that they reston a fiction, but believe that fiction to be innocuous. The other,although few people will question its basic contentions if they arestated abstractly, leads in some respects to conclusions so unwel-come that few are willing to follow it through to the end.The first gives us a sense of unlimited power to realize ourwishes, while the second leads to the insight that there are limita-tions to what we can deliberately bring about, and to the recogni-tion that some of our present hopes are delusions. Yet the effect ofallowing ourselves to be deluded by the first view has always beenthat n1an has actually limited the scope of what he can achieve. Forit has always been the recognition of the limits of the possible whichhas enabled man to make full use of his powers. 1The first view holds that human institutions will serve humanpurposes only if they have been deliberately designed for thesepurposes, often also that the fact that an institution exists is evi-dence of its having been created for a purpose, and always that we8REASON AND EVOLUTIONshould so re-design society and its institutions that all our actionswill be wholly guided by known purposes. To most people thesepropositions seem almost self-evident and to constitute an attitudealone worthy of a thinking being. Yet the belief underlying them,that we owe all beneficial institutions to design, and that only suchdesign has made or can make them useful for our purposes, islargely false.This view is rooted originally in a deeply ingrained propensity ofprimitive thought to interpret all regularity to be found in pheno-mena anthropomorphically, as the result of the design of a thinkingmind. But just when man was well on the "vay to emancipatinghimself from this naive conception, it was revived by the supportof a powerful philosophy with which the aim of freeing the humanmind from false prejudices has become closely associated, and whichbecame the dominant conception of the Age of Reason.The other view, which has slowly and gradually advanced sinceantiquity but for a time was almost entirely overwhelmed by themore glamorous constructivist view, was that that orderliness ofsociety which greatly increased the effectiveness of individual actionwas not due solely to institutions and practices which had beeninvented or designed for that purpose, but was largely due to a pro-cess described at first as 'growth' and later as 'evolution', a processin which practices which had first been adopted for other reasons,or even purely accidentally, were preserved because they enabledthe group in which they had arisen to prevail over others. Since itsfirst systematic development in the eighteenth century this viewhad to struggle not only against the anthropomorphism of primi-tive thinking but even more against the reinforcement these naiveviews had received from the new rationalist philosophy. It was in-deed the challenge which this philosophy provided that led to theexplicit formulation of the evolutionary view. 2The tenets of Cartesian rationalismThe great thinker from whom the basic ideas of what we shall callconstructivist rationalism received their most complete expressionwas Rene Descartes. But while he refrained from drawing the con-clusions from them for social and moral arguments, 3these weremainly elaborated by his slightly older (but much more long-lived)contemporary, Thomas Hobbes. Although Descartes' immediateconcern was to establish criteria for the truth of propositions, these9REASON AND EVOLUTIONwere inevitably also applied by his follo\vers to judge the appropri-ateness and justification of actions. The 'radical doubt' whichmade him refuse to accept anything as true which could not belogically derived from explicit premises that were 'clear and dis-tinct', and therefore beyond possible doubt, deprived of validity allthose rules of conduct which could not be justified in this manner.Although Descartes himself could escape the consequences byascribing such rules of conduct to the design of an omniscientdeity, for those among his followers to whom this no longer seemedan adequate explanation the acceptance of anything which wasbased merely on tradition and could not be fully justified on rationalgrounds appeared as an irrational superstition. The rejection as'mere opinion' of all that could not be demonstrated to be true byhis criteria became the dominant characteristic of the movementwhich he started.Since for Descartes reason was defined as logical deduction fromexplicit premises, rational action also came to mean only such actionas was determined entirely by known and demonstrable truth. It isalmost an inevitable step from this to the conclusion that onlywhat is true in this sense can lead to successful action, and thattherefore everything to which man owes his achievements is aproduct of his reasoning thus conceived. Institutions and practiceswhich have not been designed in this n1anner can be beneficialonly by accident. Such became the characteristic attitude ofCartesian constructivism with its contempt for tradition, custom,and history in general. Man's reason alone should enable him toconstruct society anew. 4This 'rationalist' approach, however, meant in effect a relapseinto earlier, anthropomorphic modes of thinking. It produced a re-ne\ved propensity to ascribe the origin of all institutions of cultureto invention or design. Morals, religion and law, language andwriting, money and the market, were thought of as having beendeliberately constructed by somebody, or at least as owing what-ever perfection they possessed to such design. This intentionalist orpragmatic 5 account of history found its fullest expression in theconception of the formation of society by a social contract, first inHobbes and then in Rousseau, who in many respects was a directfollo\ver of Descartes. 6 Even though their theory was not alvvaysmeant as a historical account of what actually happened, it wasalways meant to provide a guideline for deciding whether or notexisting institutions were to be approved as rational.10REASON AND EVOLUTIONI t is to this philosophical conception that we owe the preferencewhich prevails to the present day for everything that is done'consciously' or 'deliberately', and from it the terms 'irrational'or 'non-rational' derive the derogatory meaning they now have.Because of this the earlier presumption in favour of traditional orestablished institutions and usages became a presumption againstthem, and 'opinion' came to be thought of as 'mere' opinion-something not demonstrable or decidable by reason and thereforenot to be accepted as a valid ground for decision.Yet the basic assumption underlying the belief that man hasachieved n1astery of his surroundings mainly through his capacityfor logical deduction from explicit premises is factually false, andany attempt to confine his actions to what could thus be justifiedwould deprive him of many of the most effective means to successthat have been available to him. It is simply not true that ouractions owe their effectiveness solely or chiefly to knowledge whichwe can state in \vords and \vhich can therefore constitute the ex-plicit premises of a syllogism. Many of the institutions of societywhich are indispensable conditions for the successful pursuit ofour conscious aims are in fact the result of customs, habits orpractices which have been neither invented nor are observed withany such purpose in view. We live in a society in which we cansuccessfully orientate ourselves, and in which our actions have agood chance of achieving their aims, not only because our fellowsare governed by known aims or known connections between meansand ends, but because they are also confined by rules whose pur-pose or origin we often do not know and of whose very existencewe are often not aware.Man is as much a rule-following animal as a purpose-seekingone. 7 And he is successful not because he knows why he ought toobserve the rules \vhich he does observe, or is even capable ofstating all these rules in \vords, but because his thinking and actingare governed by rules which have by a process of selection beenevolved in the society in which he lives, and \vhich are thus theproduct of the experience of generations.The permanent limitations of our factual knowledgeThe constructivist approach leads to false conclusions because man'sactions are largely successful, not merely in the primitive stage butperhaps even more so in civilization, because they are adapted bothIIREASON AND EVOLUTIONto the particular facts which he knows and to a great many otherfacts he does not and cannot know. And this adaptation to thegeneral circumstances that surround him is brought about by hisobservance of rules which he has not designed and often does noteven knovv explicitly, although he is able to honour them in action.Or, to put this differently, our adaptation to our environment doesnot consist only, and perhaps not even chiefly, in an insight intothe relations between cause and effect, but also in our actions beinggoverned by rules adapted to the kind of world in which we live,that is, to circumstances which we are not aware of and which yetdetermine the pattern of our successful actions.Complete rationality of action in the Cartesian sense demandscomplete knowledge of all the relevant facts. A designer or engi-neer needs all the data and full power to control or manipulatethem if he is to organize the material objects to produce the in-tended result. But the success of action in society depends on moreparticular facts than anyone can possibly know. And our wholecivilization in consequence rests, and must rest, on our believingrnuch that we cannot knowto be true in the Cartesian sense.What we must ask the reader to keep constantly in mind through-out this book, then, is the fact of the necessary and irremediableignorance on everyone's part of most of the particular facts whichdetermine the actions of all the several members of human society.This may at first seem to be a fact so obvious and incontestableas hardly to deserve mention, and still less to require proof. Yetthe result of not constantly stressing it is that it is only too readilyforgotten. This is so mainly because it is a very inconvenient factwhich makes both our attempts to explain and our attempts toinfluence intelligently the processes of society very much moredifficult, and which places severe limits on what we can say or doabout them. There exists therefore a great temptation, as a firstapproximation, to begin with the assumption that we know every-thing needed for full explanation or control. This provisional as-sumption is often treated as something of little consequence whichcan later be dropped without much effect on the conclusions. Yetthis necessary ignorance of most of the particulars which enter theorder of a Great Society is the source of the central problem of allsocial order and the false assumption by which it is provisionallyput aside is mostly never explicitly abandoned but merely con-veniently forgotten. The argument then proceeds as if that ignor-ance did not matter.12REASON AND EVOLUTIONThe fact of our irrcrnediable ignorance of most of the particularfacts which determine the processes of society is, however, thereason why most social institutions have taken the form they actu-ally have. To talk about a society about vvhich either the observeror any of its members knows all the particular facts is to talk aboutsomething wholly different from anything \vhich has ever existcd-a society in which lnost of \vhat \ve find in our society \vould notand could not exist and \vhich, if it ever occurred, \vould possessproperties \ve cannot even imagine.I have discussed the importance of our necessary ignorance ofthe concrete facts at some length in an earlier book 8 and willemphasize its central importance here mainly by stating it at thehead of the \vhole exposition. But there are several points \vhichrequire re-statement or elaboration. In the first instance, the incur-able ignorance of everyone which I am speaking is the ignoranceof particular facts which are or will become kno\vn to somebody andthereby affect the \vhole structure of society. rrhis structure ofhuman activities constantly adapts itself, and functions throughadapting itself, to millions of facts which in their entirety are notknown to anybody. The significance of this process is most obviousand \\Tas at first stressed in the economic field. As it has been said,'the economic life of a non-socialist society consists of millions ofrelations or flows between individual firms and households. \Vecan establish certain theorems about them, but vve can neverobserve all.' 9 The insight into the significance of our institutionalignorance in the economic sphere, and into the methods by \vhich\ve have learnt to overcome this obstacle, \vas in fact the startingpoint 10 for those ideas which in the present book arc systelnaticallyapplied to a much wider field. It will be one of our chief contentionsthat most of the rules of conduct \vhich govern our actions, andlnost of the institutions which arise out of this regularity, areadaptations to the impossibility of anyone taking conscious accountof all the particular facts which enter into the order of society.vVe shall see, in particular, that the possibility of justice rests on thisnecessary limitation of our factual knowledge, and that insight intothe nature of justice is therefore denied to all those constructivists\\ho habitually argue on the assulnption of omniscience.Another consequence of this basic fact \vhich must be stressedhere is that only in the small groups of primitive society cancollaboration bet\veen the members rest largely on the circumstancethat at anyone moment theywill knowmore or less the same particular13REASON AND EVOLUTIONcirculnstances. SOl1le wise men 111ay be better at interpretingthe immediately perceived circumstances or at remembering thingsin rClnote places unkndvvn to the others. But the concrete events\vhich the individuals encounter in their daily pursuits will be verymuch the same for all, and they will act together because the eventsthey know and the objectives at which they aim are more or lessthe same.The situation is wholly different in the Great 11 or Open Societywhere millions of men interact and where civilization as we know ithas developed. Econon1ics has long stressed the 'division of labour'which such a situation involves. But it has laid much less stress onthe fragmentation of knowledge, on the fact that each Inember ofsociety can have only a small fraction of the knowledge possessedby all, and that each is therefore ignorant of most of the facts onwhich the working of society rests. Yet it is the utilization of muchmore knowledge than anyone can possess, and therefore the factthat each moves within a coherent structure most of whose deterlni-nants are unknown to him, that constitutes the distinctive featureof all advanced civilizations.In civilized society it is indeed not so much the greater know-ledge that the individual can acquire, as the greater benefit he re-ceives from the kno\vledge possessed by others, which is the causeof his ability to pursue an infinitely wider range of ends than merelythe satisfaction of his most pressing physical needs. Indeed, a'civilized' individual may be very ignorant, more ignorant thanmany a savage, and yet greatly benefit from the civilization inwhich he lives.The characteristic error of the constructivist rationalists in thisrespect is that they tend to base their argument on what has beencalled the synoptic delusion, that is, on the fiction that all the rele-vant facts are known to some one mind, and that it is possible toconstruct from this knowledge of the particulars a desirable socialorder. Sometimes the delusion is expressed with a touching naiveteby the enthusiasts for a deliberately planned society, as when oneof them dreams of the development of 'the art of simultaneousthinking: the ability to deal with a multitude of related phenomenaat the same time, and of composing in a single picture both thequalitative and the quantitative attributes of these phenomena.' 12They seem completely unaware that this dream simply assumesaway the central problem which any effort towards the understand-ing or shaping of the order of society raises: our incapacity toREASON AND EVOLUTIONassemble as a surveyable whole all the data \vhich enter into thesocial order. Yet all those \vho are fascinated by the beautifulplans which result from such an approach because they are 'soorderly, so visible, so easy to understand', 13 are the victims of thesynoptic delusion and forget that these plans o\ve their seemingclarity to the planner's disregard of all the facts he does notknow.Factual knowledge and scienceThe chief reason why n10dern man has become so unwilling toadmit that the constitutional limitations on his knovvledge form apermanent barrier to the possibility of a rational construction of thewhole of society is his unbounded confidence in the ofscience. We hear so much about the rapid advance of scientifickno\vledge that we have come to feel that all luere lin1itations ofkno\vledge are soon bound to disappear. confidence rests,ho\vever, on a misconception of the tasks and powers of science,that is, on the erroneous belief that science is a method of ascer-taining particular facts and that the progress of its techniques \villenable us to ascertain and manipulate all the particular facts wemight want.In one sense the saying that our civilization rests on the con-quest of ignorance is of course a mere platitude. Yet our veryfamiliarity \\lith it tends to conceal from us \vhat is most ilnportantin it: namely that civilization rests on the fact that \VC all benefitfrom knowledge which we do not possess. And one of the ways inwhich civilization helps us to overcome that limitation on the ex-tent of individual knowledge is by conquering ignorance, not bythe acquisition of more knowledge, but by the utilization of know-ledge which is and remains widely dispersed alnong individuals.The limitation of knowledge with which we arc concerned is there-fore not a limitation which science can overcome. Contrary to awidely held belief, science consists not of the kno\\'ledge of particu-lar facts; and in the case of very complex phenoluena the po\vers ofscience are also limited by the practical impossibility of ascertainingall the particular facts which we would have to know if its theories\-vere to give us the po\ver of predicting specific events. The studyof the relatively simple phenomena of the physical world, \vhere ithas proved possible to state the determining relations as functionsof a fe\v variables that can be easily ascertained in particularREASON AND EVOLUTIONinstances, and where as a consequence the astounding progress ofdisciplines concerned with them has become possible, has createdthe illusion that soon the same will also be true with regard to themore complex phenomena. But neither science nor any knowntechnique 14 enables us to overcome the fact that no mind, andtherefore also no deliberately directed action, can take account of allthe particular facts which are known to some men but not as a\vhole to any particular person.Indeed, in its endeavour to explain and predict particular events,which it does so successfully in the case of relatively simple phe-nomena (or where it can at least approximately isolate 'closedsystems' that are relatively simple), science encounters the samebarrier of factual ignorance when it comes to apply its theories tovery complex phenomena. In some fields it has developed import-ant theories which give us much insight into the general characterof some phenomena, but will never produce predictions of particu-lar events, or a full explanation-simply because we can neverknow all the particular facts \vhich according to these theories wewould have to know in order to arrive at such concrete conclusions.The best example of this is the Darwinian (or Neo-Darwinian)theory of the evolution of biological organisms. If it were possibleto ascertain the particular facts of the past which operated on theselection of the particular forms that emerged, it would provide acomplete explanation of the structure of the existing organisms;and similarly, if it were possible to ascertain all the particular factswhich will operate on them during some future period, it ought toenable us to predict future development. But, of course, we willnever be able to do either, because science has no means of ascer-taining all the particular facts that it would have to possess to per-form such a feat.There is another related misconception about the aim andpower of science which it will be useful also to mention at this point.This is the belief that science is concerned exclusively with whatexists and not with what could be. But the value of science con-sists largely in telling us what would happen if some facts weredifferent from what they are. All the statements of theoreticalscience have the form of 'if ... , then ...' statements, andthey are interesting mainly in so far as the conditions we insert inthe 'if' clause are different from those that actually exist.Perhaps this misconception has nowhere else been so importantas in political science where it seems to have become a bar to16REASON AND EVOLUTIONserious consideration of the really important problems. Here themistaken idea that science is simply a collection of observed factshas led to a confinement of research to the ascertainment of what is,vvhile the chief value of all science is to tell us what the conse-quences would be if conditions were in some respects made differ-ent from what they are.The fact that an increasing number of social scientists confinethemselves to the study of what exists in some part of the socialsystem does not make their results more realistic, but makes themlargely irrelevant for most decisions about the future. Fruitfulsocial science must be very largely a study of what is not: a con-struction of hypothetical models of possible worlds which n1ightexist if some of the alterable conditions were made different. vVeneed a scientific theory chiefly to tell us \vhat would be the effects ifsome conditions were as they have never been before. All scientificknowledge is knowledge not of particular facts but of hypotheseswhich have so far withstood systematic attempts at refuting them.The concurrent evolution ofmind and society: the role ofrulesThe errors of constructivist rationalism are closely connected withCartesian dualism, that is with the conception of an independentlyexisting mind substance which stands outside the cosmos of natureand which enabled man, endowed with such a mind from the be-ginning, to design the institutions of society and culture amongwhich he lives. The fact is, of course, that this mind is an adaptationto the natural and social surroundings in which man lives and thatit has developed in constant interaction with the institutions whichdetermine the structure of society. Mind is as much the product ofthe social environlnent in which it has grown up and which it hasnot made as something that has in turn acted upon and alteredthese institutions. It is the result of man having developed in soci-ety and having acquired those habits and practices that increasedthe chances of persistence of the group in which he lived. Theconception of an already fully developed mind designing the insti-tutions which made life in society possible is contrary to all weknow about the evolution of man.The cultural heritage into which man is born consists of a com-plex of practices or rules of conduct which have prevailed becausethey made a group of men successful but which were not adoptedbecause it was known that they would bring about desired effects.REASON AND EVOLUTIONMan acted before he thought and did not understand before heacted. What we call understanding is in the last resort simply hiscapacity to respond to his environment \vith a pattern of actionsthat helps him to persist. Such is the modicum of truth in behavi-ourism and pragmatism, doctrines which, however, have so crudelyoversimplified the determining relationships as to becolne moreobstacles than helps to their appreciation.'Learning from experience', among men no less than amonganimals, is a process not primarily of reasoning but of the obser-vance, spreading, transmission and development of practices whichhave prevailed because they were successful-often not becausethey conferred any recognizable benefit on the acting individual butbecause they increased the chances of survival of the group to whichhe belonged. 15The result of this development will in the firstinstance not be articulated knowledge but a knowledge which,although it can be described in terms of rules, the individual can-not state in words but is merely able to honour in practice. Themind does not so much make rules as consist of rules of action, acomplex of rules that is, which it has not made, but which havecome to govern the actions of the individuals because actions inaccordance with thCIll have proved more successful than those ofcompeting individuals or groups. 16There is in the beginning no distinction between the practicesone must observe in order to achieve a particular result and thepractices one ought to observe. There is just one established man-ner of doing things, and knowledge of cause and effect and know-ledge of the appropriate or permissible form of action are notdistinct. I(nowledge of the world is knowledge of what one must door not do in certain kinds of circulnstances. And in avoiding dangerit is as important to know what one must never do as to know whatone must do to achieve a particular result.These rules of conduct have thus not developed as the recog-nized conditions for the achievement of a known purpose, but haveevolved because the groups who practised them \vere more suc-cessful and displaced others. They were rules which, given the kindof environment in which man lived, secured that a greater numberof the groups or individuals practising them would survive. Theproblem of conducting himself successfully in a world only parti-ally known to n1an \vas thus solved by adhering to rules which hadserved him well but which he did not and could not know to betrue in the Cartesian sense.18REASON AND EVOLUTIONThere are thus two attributes of these rules that govern humanconduct and make it appear intelligent which we shall have tostress throughout, because the constructivist approach denies im-plicitly that it can be rational to observe such rules. Of course inadvanced society only SOlne rules \vill be of this kind; \vhat \vewant to emphasize is merely that even such advanced societies,"rill in part owe their order to son1e such rules.The first of these attributes which most rules of conduct origi-nally possessed is that they are observed in action without beingkno\vn to the acting person in articulated ('verbalized' or explicit)form. They " , ~ i l l manifest themselves in a regularity of action \vhichcan be explicitly described, but this regularity of action is not theresult of the acting persons being capable of thus stating them. r-rhesecond is that such rules con1C to be observed because in fact theygive the group in which they are practised superior strength, andnot because this effect is known to those who are guided by then1.Although such rules come to be generally accepted because theirobservation produces certain consequences, they are not observedwith the intention of producing those consequences-consequenceswhich the acting person need not know.We cannot consider here the difficult question of how men canlearn froln each other such, often highly abstract, rules of conductby exalnple and imitation (or 'by analogy'), although neither those\vho set the examples nor those \vho learn from them n1ay be con-sciously a\vare of the existence of the rules ",-hich they neverthelessstrictly observe. This is a problem n10st fan1iliar to us in the learn-ing of language by children who are able to produce correctly lnostcomplicated expressions they have never heard before; 17 but itoccurs also in such fields as Juanners, n10rals and la\v, and in mostskills where we are guided by rules \vhich \ve kno\v ho\v to followbut are unable to state.The ilnportant point is that every n1an gro,"ring up in a givenculture will find in himself rules, or may discover that he acts inaccordance with rules-and will similarly recognize the actions ofothers as conforming or not conforming to various rules. This is,of course, not proof that they are a permanent or unalterable part of'hulnan nature', or that they are innate, but proof only that they arepart of a cultural heritage which is likely to be fairly constant,especially so long as they are not articulated in \vords and thereforealso are not discussed or consciously examined.REASON AND EVOLUTIONJ'he false dichotomy of 'natural' and 'artificial'The discussion of the problems with which we are concerned waslong hampered by the universal acceptance of a tnisleading distinc-tion which was introduced by the ancient Greeks and from whoseconfusing effect we have not yet wholly freed ourselves. This is thedivision of phenomena between those which in modern terms are'natural' and those which are 'artificial'. The original Greek terms,which seem to have been introduced by the Sophists of the fifthcentury B.C., were physei, which means 'by nature' and, in contrastto it, either nomo, best rendered as 'by convention', or thesei,which means roughly 'by deliberate decision'. 18 The use of twoterms with somewhat different meanings to express the second partof the division indicates the confusion which has beset the dis-cussion ever since. The distinction intended may be either betweenobjects which existed independently and objects which were theresults of human action, or between objects which arose indepen-dently of, and objects which arose as the result of, human design.The failure to distinguish bet\veen these two meanings led to thesituation where one author could argue with regard to a givenphenomenon that it was artificial because it was the result of humanaction, while another might describe the same phenomenon asnatural because it was evidently not the result of human design.Not until the eighteenth century did thinkers like Bernard Mande-ville and David Hume make it clear that there existed a category ofphenomena which, depending on which of the two definitions oneadhered to, would fall into either the one or the other of the twocategories and therefore ought to be assigned to a distinct third classof phenomena, later described by Adam Ferguson as 'the result ofhuman action but not of human design' .19 These were the phe-nomena which required for their explanation a distinct body oftheory and \vhich came to provide the object of the theoreticalsocial sciences.But in the more than two thousand years during which the dis-tinction introduced by the ancient Greeks has ruled thought almostunchallenged, it has become deeply engrained in concepts andlanguage. In the. second century A.D. a Latin grammarian, AulusGellius, rendered the Greek terms physei and thesei by naturalisand positivus, from which most European languages derived thewords to describe two kinds of law. 20There occurred later one promising development in the dis-20REASON AND EVOLUTIONcussion of these questions by the medieval schoolmen, which ledclose to a recognition of the intermediate category of phenomenathat were 'the result of hun1an action but not of human design'. Inthe twelfth century some of those writers had begun to include undernaturalis all that was not the result of human invention or a deliber-ate creation; 21 and in the course of time it came to be increasinglyrecognized that many social phenomena fell into this category.Indeed, in the discussion of the problems of society by the last ofthe schoolmen, the Spanish Jesuits of the sixteenth century, natura-lis became a technical term for such social phenomena as were notdeliberately shaped by human will. In the work of one of them,Luis Molina, it is, for example, explained that the 'natural price' isso called because 'it results from the thing itself without regard tolaws and decrees, but is dependent on many circumstances whichalter it, such as the sentiments of men, their estimation of differentuses, often even in consequence of whims and pleasures'. 22 In-deed, these ancestors of ours thought and 'acted under a strong im-pression of the ignorance and fallibility of mankind', 23 and, forinstance, argued that the precise 'mathematical price' at which acommodity could be justly sold was only known to God, because itdepended on more circumstances than any man could know, andthat therefore the determination of the 'just price' must be leftto the market. 24These beginnings of an evolutionary approach were sub-merged, however, in the sixteenth and seventeenth centuries bythe rise of constructivist rationalism, with the result that both theterm 'reason' and the term 'natural law' completely changed theirmeaning. 'Reason', which had included the capacity of the mind todistinguish between good and evil, that is between what was andwhat was not in accordance with established rules, 25 came to meana capacity to construct such rules by deduction from explicit premi-ses. The conception of natural law was thereby turned into that of a'law of reason' and thus almost into the opposite of what it hadmeant. This new rationalist law of nature of Grotius and his suc-cessors, 26 indeed, shared with its positivist antagonists the concep-tion that all law was made by reason or could at least be fullyjustified by it, and differed from it only in the assumption that lawcould be logically derived from a priori premises, while positivismregarded it as a deliberate construction based on empirical know-ledge of the effects it would have on the achievement of desirablehuman purposes.21REASON AND EVOLUTIONThe rise of the evolutionary approachAfter the Cartesian relapse into anthropomorphic thinking on thesematters a new start was made by Bernard Mandeville and DavidHume. They were probably inspired more by the tradition of theEnglish common law, especially as expounded by Matthew Hale,than by the the law of nature. 27 It came increasingly to be seenthat the formation of regular patterns in human relations that werenot the conscious aim of human actions raised a problem which re-quired the development of a systematic social theory. This needwas met during the second half of the eighteenth century in thefield of economics by the Scottish moral philosophers, led by AdamSmith and Adam Ferguson, while the consequences to be drawnfor political theory received their magnificent formulations fromthe great seer Edmund Burke, in whose \\lork we shall, ho\vever,seek in vain for