Contents

Cover

About the Book

About the Author

Also by Isaiah Berlin

Title Page

Editor’s Preface

Author’s Preface

Introduction by Bernard Williams

The Purpose of Philosophy

Verification

Empirical Propositions and Hypothetical Statements

Logical Translation

Equality

The Concept of Scientific History

Does Political Theory Still Exist?

‘From Hope and Fear Set Free’

Index

Copyright

About the Book

Although Isaiah Berlin liked to say that he left philosophy for the history of ideas after the Second World War, there is a decided continuity between his more purely philosophical writings, most of which are collected in this volume, and the more historical work for which he is better known. Included here are Berlin’s early arguments against logical positivism and later essays which more evidently reflect his life-long interest in political theory, intellectual history and the philosophy of history. In two related pieces he gives his view on the philosopher’s task, to uncover the various models – the concepts and categories – that we bring to our experience, and that help to form it. In his own words: ‘The goal of philosophy is always the same, to assist men to understand themselves and thus operate in the open, and not wildly, in the dark.’

About the Author

Sir Isaiah Berlin, OM, who died in 1997, was born in Riga, capital of Latvia, in 1909. When he was six, his family moved to Russia: there in 1917, in Petrograd, he witnessed both the Social-Democratic and the Bolshevik Revolutions.

In 1921 his family came to England, and he was educated at St Paul’s School and Corpus Christi College, Oxford. At Oxford he was a Fellow of All Souls, a Fellow of New College, Professor of Social and Political Theory, and founding President of Wolfson College. He also held the Presidency of the British Academy. His published work includes Karl Marx, Four Essays on Liberty, Vico and Herder, Russian Thinkers, Against the Current, Personal Impressions, The Crooked Timber of Humanity, The Magus of the North, The Sense of Reality, The Proper Study of Mankind and The Roots of Romanticism. As an exponent of the history of ideas he was awarded the Erasmus, Lippincott and Agnelli Prizes; he also received the Jerusalem Prize for his lifelong defence of civil liberties.

Dr Henry Hardy, a Supernumerary Fellow of Wolfson College, Oxford, is one of Isaiah Berlin’s literary trustees. He has edited several other collections of Berlin’s work, and is currently preparing his remaining unpublished writings and his letters for publication.

Also by Isaiah Berlin

KARL MARX

THE AGE OF ENLIGHTENMENT

FOUR ESSAYS ON LIBERTY

VICO AND HERDER

Edited by Henry Hardy and Aileen Kelly

RUSSIAN THINKERS

Edited by Henry Hardy

AGAINST THE CURRENT

PERSONAL IMPRESSIONS

THE CROOKED TIMBER OF HUMANITY

THE MAGUS OF THE NORTH

THE SENSE OF REALITY

THE ROOTS OF ROMANTICISM

Edited by Henry Hardy and Roger Hausheer

THE PROPER STUDY OF MANKIND

CONCEPTS AND CATEGORIES

Philosophical Essays

ISAIAH BERLIN

Edited by Henry Hardy

With an introduction by Bernard Williams

Editor’s Preface

This is one of five volumes published in Isaiah Berlin’s lifetime in which I brought together, and prepared for reissue, most of his published essays that had not previously been made available in a collected form.¹ Until then his many writings had been scattered, often in obscure places; most were out of print; and only half a dozen essays had been collected and reissued.² These five volumes, together with the list of his publications that one of them (Against the Current) contains,³ and two subsequent volumes in which I published much of his previously unpublished work,⁴ made much more of his oeuvre readily accessible than before. Nevertheless, a good deal still remains unpublished, because Berlin felt that it needed revision, and he did not want it to appear in unrevised form in his lifetime. However, he placed no restrictions on posthumous publication, and I hope to make much of this material, which is of great interest and value, available in years to come. Indeed, the first instalment, Berlin’s 1965 Mellon Lectures on romanticism,⁵ is published at the same time as this reissue of Concepts and Categories.

The essays in the present volume are Berlin’s published contributions to philosophy, with the exception of ‘Historical Inevitability’ and ‘Two Concepts of Liberty’, which had already been reissued in revised form in Four Essays on Liberty, and ‘Induction and Hypothesis’,⁶ which is the second contribution to a symposium, and, being in large part a reply to the first contribution, does not stand comfortably on its own, though it does not deserve to be neglected by students of Berlin’s philosophical work.⁷ ‘The Purpose of Philosophy’ first appeared in Insight (Nigeria) 1 No 1 (July 1962), and was reprinted in the Sunday Times, 4 November 1962, and as ‘Philosophy’s Goal’ in Leonard Russell (ed.), Encore, 2nd Year (London, 1963: Michael Joseph); ‘Verification’, ‘Logical Translation’, ‘Equality’ and ‘“From Hope and Fear Set Free”’ (the Presidential Address for the 1963–4 Session) all appeared in the Proceedings of the Aristotelian Society, vols 39 (1938–9), 50 (1949–50), 56 (1955–6) and 64 (1963–4) respectively; ‘Empirical Propositions and Hypothetical Statements’ was published in Mind 59 (1950); ‘The Concept of Scientific History’ appeared as ‘History and Theory: The Concept of Scientific History’ in History and Theory 1 (1960) and in Alexander V. Riasanovsky and Barnes Riznik (eds), Generalizations in Historical Writing (Philadelphia, 1963: University of Pennsylvania Press), and under its present title in William H. Dray (ed.), Philosophical Analysis and History (New York, 1966: Harper and Row); ‘Does Political Theory Still Exist?’ was published first in French as ‘La théorie politique existe-t-elle?’ in Revue française de science politique 11 (1961), and then in English in Peter Laslett and W. G. Runciman (eds), Philosophy, Politics and Society, 2nd Series (Oxford, 1962: Blackwell). Apart from necessary corrections and the addition of missing references, the essays appear here essentially in their original form.

I am greatly indebted to Bernard Williams, who not only wrote the introduction to this volume, but also played a crucial role in persuading Berlin that these philosophical essays were well worth reprinting. Without his support this volume would not have appeared. Berlin himself was unfailingly courteous, good-humoured and informative in response both to my persistent general advocacy of the whole project, which he continued to regard with considerable scepticism, especially in the case of the present volume, and to my often over-meticulous probings into points of detail. Pat Utechin, his secretary, was an indispensable source of help and encouragement at all stages, and Kate McKenzie kindly double-checked the proofs.

HENRY HARDY

May 1998

¹ This volume was first published in London in 1978, and in New York in 1979. The other volumes are Russian Thinkers (London and New York, 1978), co-edited with Aileen Kelly; Against the Current: Essays in the History of Ideas (London, 1979; New York, 1980); Personal Impressions (London, 1980; New York, 1981; 2nd ed., London, 1998); and The Crooked Timber of Humanity: Chapters in the History of Ideas (London, 1990; New York, 1991). There is also a one-volume selection of essays drawn from these collections and their predecessors: The Proper Study of Mankind: An Anthology of Essays, ed. Henry Hardy and Roger Hausheer (London, 1997: New York, 1998).

² Four Essays on Liberty (London, 1969; New York, 1970) and Vico and Herder: Two Studies in the History of Ideas (London and New York, 1976). Other collections had appeared only in translation.

³ Its currently most up-to-date version appears in the Pimlico paperback edition published in London in 1997.

⁴ The Magus of the North: J. G. Hamann and the Origins of Modern Irrationalism (London, 1993; New York, 1994) and The Sense of Reality: Studies in Ideas and their History (London, 1996; New York, 1997).

⁵ The Roots of Romanticism (London and Princeton, 1999).

⁶ Proceedings of the Aristotelian Society, supplementary vol. 16 (1937).

⁷ Readers may like to have details of four other pieces in this area which do not appear here. There is a long review of Bertrand Russell’s A History of Western Philosophy in Mind 56 (1947); ‘Philosophy and Beliefs’, Twentieth Century 157 (1955), is a conversation with Anthony Quinton, Stuart Hampshire and Iris Murdoch; ‘An Introduction to Philosophy’ is an interview with Bryan Magee in Men of Ideas (London, 1978); and ‘Is a Philosophy of History Possible?’ is a discussion with others in Yirmiahu Yovel (ed.), Philosophy of History and Action (Dordrecht/Boston/London and Jerusalem, 1978). For other reviews and smaller pieces, see the bibliography already cited.

Author’s Preface

Some of these articles were written more than a quarter of a century ago, when I was teaching philosophy in Oxford; when Dr Henry Hardy proposed to me that they should be included in a separate volume, I demurred. Although I do not think that there is anything in them that I should now wish to withdraw or change radically (I could not bring myself to re-read them), it seemed to me that they belonged too much to their time and place – they were not untypical of the kind of discussions and controversies, mainly about positivism, that went on in Oxford in the years immediately before and after the war, but I thought that they contained little or nothing worth resuscitating nearly thirty years later. I felt similar doubts about the articles written in the years that followed. Dr Hardy thought better of these pieces than I did, and when I continued to be obdurate, he proposed that we should go to arbitration and suggested that Professor Bernard Williams be appealed to. Bernard Williams is an original philosopher and a just and candid critic, and I therefore expected him to agree with me. When he said that he favoured republication, I could not, of course, help being pleased, and I accepted his verdict even though I wondered whether it was not more generous than just. Dr Hardy pressed his advantage and persuaded Professor Williams to back his judgement by writing an introduction to the volume. For this act of what I can only describe as heroic friendship I record my deep gratitude.

I have occasionally been asked what made me cease to teach philosophy as it is taught in most English-speaking universities, and as I believe it should be taught. The answer is best given by recording a conversation I had with the late Professor H. M. Sheffer of Harvard, whom I met there towards the end of the war when I was working at the British Embassy in Washington. Sheffer, one of the most eminent mathematical logicians of his day, said to me that in his opinion there were only two philosophical disciplines in which one could hope for an increase of permanent knowledge: one was logic, in which new discoveries and techniques superseded the old ones – this was a field of exact knowledge in which genuine progress occurred, as it did in the natural sciences or mathematics; the other was psychology, which he thought of as being in some respects still philosophical – this was an empirical study and obviously capable of steady development. And, of course, there was the history of philosophy: but this was not part of philosophy itself; as for logic and psychology, they differed from philosophy proper, to which – unlike history or classical learning – the notion of growth, of cumulative knowledge, did not seem to him to apply. ‘To speak of a man learned in epistemology, or a scholar in ethics,’ he said, ‘does not make sense; it is not that kind of study.’ He went on to say that philosophy was a marvellous province of thought, but it had not been helped, in his view, indeed had been gravely damaged, by what logical positivists, influenced by symbolic logicians like himself, were now doing; the kind of work that ‘Carnap and Co.’ (as he called them) were engaged upon repelled him – it would ruin real philosophy as he and his master Royce conceived it: ‘If any work of mine has done anything to stimulate this development, I had rather not have been born.’ Although I did not, and do not, agree with Sheffer’s sweeping condemnation of the value and influence of logical positivism, or the rigid division he drew, repudiating his own earlier views, between logic and philosophy, his words made a profound impression upon me. In the months that followed, I asked myself whether I wished to devote the rest of my life to a study, however fascinating and important in itself, which, transforming as its achievements undoubtedly were, would not, any more than criticism or poetry, add to the store of positive human knowledge. I gradually came to the conclusion that I should prefer a field in which one could hope to know more at the end of one’s life than when one had begun; and so I left philosophy for the field of the history of ideas, which had for many years been of absorbing interest to me.

My reason for telling this story is mainly historical, because of the light it throws on the conception of philosophy held towards the end of his life by one of the fathers of modern logic, about whose general views little or nothing, so far as I know, has been published; and also because somewhat inaccurate accounts of this conversation have been in circulation, one of which has recently found its way into print – and I thought it as well to set the record straight.

ISAIAH BERLIN

February 1978

Introduction

Bernard Williams

Isaiah Berlin is most widely known for his writings in political theory and the history of ideas, but he worked first in general philosophy, and contributed to the discussion of those issues in the theory of knowledge and the theory of meaning which preoccupied the more radical among the young philosophers at Oxford in the late 1930s. The medium was in good part personal discussion, particularly within a group including Stuart Hampshire, the late A. J. Ayer, the late J. L. Austin, and others.¹ In this selection from Berlin’s more purely philosophical writings, the three papers which represent that earliest period of his concerns (only one of them written actually pre-war) involve the reader in a double displacement from what those philosophical conversations must have been like. The transition from dialectic to document is one thing – something that many philosophers of many schools have found problematical. Another thing is the transition from Berlin in person to Berlin in print.

It has been said that the kind of philosophical activity engaged in then by Berlin and his friends, like the ‘linguistic philosophy’ of the 40s and 50s which it helped to form, was essentially conversational and resisted publication. So far as the real point of the activity was concerned, as opposed to a certain manner, this has probably been exaggerated. Among ‘analytical’ or ‘linguistic’ philosophers, only Wittgenstein had an understanding of the nature of philosophy which (like that of Socrates) meant that something essential to the subject itself was lost in the transition to print. Nothing that Austin (for one) believed about the subject would have precluded him from writing a textbook, even, and Ayer has not declined to do so. The present papers can, and surely do, preserve the point of those philosophical enquiries.

The second transition, however, from Berlin in discussion to Berlin in print, involves losses which are clear and determinate, even if they are hard to describe. The decorum of a journal article must attenuate that sense, which Berlin uniquely conveys, that no abstract or analytical point exists out of all connection with historical, personal, thought: that every thought belongs, not just somewhere, but to someone, and is at home in a context of other thoughts, a context which is not purely formally prescribed. Thoughts are present to Berlin not just, or primarily, as systematic possibilities, but as historically and psychologically actual, and as something to be known and understood in these concrete terms. This is one thing, besides a courteous nature, that makes Berlin a less than ruthless controversialist – a highly developed sensibility for what it is to be the other party, to see the world in that different way.

The agenda of philosophy for the group to which Berlin belonged before the war was set, in some part, by logical positivism. They were concerned with the conditions of sentences having a meaning, and with the connections between meaning and verification, where verification was construed in terms of sense-perception. Positivism both regarded natural science as the paradigm of knowledge, and took a strictly empiricist view of science, seeing scientific theory in operationalist terms as a mere compendium and generator of actual and possible observations. This set of ideas does not leave very much room for the historical imagination, nor for insight. It is hardly surprising that Berlin was never a positivist. But, seriously interested in philosophy at a time when philosophy’s most pressing questions came from a positivist direction, he produced work which did not merely reject positivism programmatically, but argued its issues in its own kind of terms. Two essays in the present book are of this kind: ‘Verification’, and ‘Empirical Propositions and Hypothetical Statements’. Both express a deep resistance to the operationalist ideas of positivism, which held that the meaning of our statements about reality is given directly by our procedures for finding out about it. Against this general conception, Berlin affirms that our understanding of reality already includes the conception of it as existing independently of us and our understanding; so that our reflection on what we mean when we characterise that reality cannot accommodate the positivist idea that truths about reality should be equivalent to truths about us.

This unacceptably idealist equivalence, as Berlin detects it to be, gets no better (as the positivists hoped) if categorical truths about reality are treated as equivalent to hypothetical truths about us (or about other possible observers). This was the manoeuvre of phenomenalism, which was par excellence the positivist theory of the external world. Phenomenalism tried to analyse all statements about the material world into statements about actual or possible experiences. Statements about observed objects were, under analysis, at least partly categorical: they recorded the actual observations. Statements about unobserved objects, on the other hand, were, when analysed, entirely hypothetical. But this conjunction of claims, as Berlin points out, cannot possibly be correct: the difference between what happens to be observed and what remains unobserved cannot possibly issue in a difference of logical form.

So, more generally, when Berlin takes up the question of a proposition’s referring to an object presently unobserved, his line of argument can be seen as striking at the mixture of epistemology and logic which has marked the empiricist tradition. (The eventual consequences of rejecting the empiricist’s epistemological notions of reference are radical, and are at the present time a major preoccupation of the philosophy of language.) One further thing that particularly comes across from Berlin’s opposition to verificationism is a powerful sense (not shared by all philosophers) of the reality of the past, something which his metaphysical opinions join the whole body of his work in affirming.

Berlin did not accept positivism’s view of meaning and knowledge, nor – above all – its view of philosophy itself as having the modest roles, up to its final retirement, of secretary to science and obituarist of metaphysics. His historical sense made him sceptical even of the more generous conception of philosophy held by post-positivist linguistic philosophy, which gave it the open-ended task of carefully and imaginatively charting the uses and implications of ordinary language, and diagnosing in those terms the origins of philosophical perplexity. Berlin claims, in ‘The Purpose of Philosophy’ and again in ‘Does Political Theory Still Exist?’, a larger task for it, in terms of an account, more perhaps in the spirit of Collingwood than of any analytical philosopher, of various models or presuppositions which men have brought to their experience, and which have helped, indeed, to form that experience. The understanding of these models, and the self-understanding of our own, are offered as one task of philosophy, and they imply others: for if the story of these various models gives a correct account of the constitution of human experience in different eras and cultural situations, then there are genuine questions about the objectivity of what is, at any given time or place, regarded as knowledge. The questions are not new, and have been explicit and pressing since (at the latest) Hegel. Linguistic philosophy had not much to say about questions of that sort and turned to other things; but the questions did not go away, or even change very much while neglected.

They can be pressed, in fact, even against natural science. Berlin has not himself done so, and indeed the one thing in these two essays that bears something of a positivist stamp is the account of science implied by his division between questions that are determinately answerable and those that are not, and the division, again, of the answerable questions into the empirical and the formal. But the activity of paradigms and models which Berlin invokes outside these domesticated areas can be detected in the development of natural science itself, as many present philosophers of science insist. Some of these philosophers, significantly, are committed to believing about scientific theories that they cannot properly be understood except in terms of their history – something which Berlin himself believes about anything that he finds really interesting.

Berlin himself has applied his concern with the role of models and presuppositions rather to the human sciences, insisting also, in ‘The Concept of Scientific History’ and elsewhere, on the peculiarity of those sciences in having a subject-matter which is of the same nature as the investigator. This feature of them, in Berlin’s view, both permits and requires from the investigator a special insightful kind of understanding, not applicable to any other kind of subject-matter. This is of course the capacity which he salutes in those – Vico and Herder first among them – who have insisted that past ages, remote cultures, saw the world through different eyes from us and that an effort of identification is needed if their view is to be in any way recaptured. It is also a capacity which Berlin himself notably displays. It applies not only to understanding across time, but also to the very different outlooks, structures of understanding and preconception, which different kinds of thinkers can bring to the world in the same period.

These various structures or models, whether across time or contemporary, inevitably raise problems of relativism: whether there is any basis on which one such view can be seen as better, more adequate, in any absolute sense, than another. Berlin offers, so far as I know, no general theoretical critique of relativism, but he is certainly resistant to it – and he has a special reason to be so, in so far as his own account of human action and its intelligibility itself implies the falsehood of some ideologies and models of life which have been influential in the past and still remain so. In ‘“From Hope and Fear Set Free”’ Berlin examines metaphysical questions about human freedom (questions which come before those issues of social and political freedom which he has discussed elsewhere), in connection with a very interesting and searching question, whether knowledge always liberates. He wants to stress the vast effect there would be on ordinary notions of action, purpose, praise, blame, regret, and so forth, if we really believed in a deterministic theory to the effect that our actions are the strict causal product of earlier states of affairs, stretching indefinitely back. The ‘reconciling’ hypothesis of self-determination, that we are free if among the causes of our action is our own choice, even though that choice itself be caused, Berlin joins Epicurus in finding not good enough, a form of ‘semi-slavery’. Berlin does not himself argue directly against determinism, nor is his denial of the reconciling strategy, his insistence that the conceptual and moral costs of believing in determinism would be enormous, intended as an argument in terrorem against accepting determinism. But the principle of self-determination he sees as definitely mistaken, and the images of liberation that go with it, to that extent flawed: absolutely flawed, not merely relatively to another set of presuppositions. Indeed one suspects that he not only hopes but believes that determinism is false, and that the whole loaf of anti-determinist freedom which the libertarian craves is actually available.

In the account that he gives of philosophy, more than one sort of question is excluded from the realm of the determinately answerable. Among them are questions of value; and the fact that they should be so excluded, and that they should be, in that context, partly assimilated to questions of philosophy, are both facts characteristic of Berlin’s outlook. That questions of value should be partly assimilated to questions of philosophy reminds us of the broad scope that Berlin gives philosophy. It also warns us that the reason why value questions are in his view ultimately contestable is not that they are ‘subjective’, or that their answers are merely expressions of opposed attitudes. Indeed, to read Berlin’s discussions of conflicts between values in the context of a debate about subjectivism is to mislocate them and to miss their special force. The debate about subjectivism is characteristically concerned with conflicts of values between persons or societies (‘Who is right?’). What above all concerns Berlin, on the other hand, is the tension between conflicting values in one consciousness.

Again and again, in these essays and elsewhere, Berlin warns us against the deep error of supposing that all goods, all virtues, all ideals are compatible, and that what is desirable can ultimately be united into a harmonious whole without loss. This is not the platitude that in an imperfect world not all the things we recognise as good are in practice compatible. It is rather that we have no coherent conception of a world without loss, that goods conflict by their very nature, and that there can be no incontestable scheme for harmonising them. There can, of course, be errors or limitations in thinking about values, whether in the particular case or in a more systematic way. For one thing, there can be the errors of omission and simplification, of succumbing to the illusion that one value can override all others and restructure everything. For Berlin, this is certainly a cardinal error, and it is in a special sense an absolute one – for it offends against something that is absolutely true about values. Yet the historical picture which Berlin also offers, the account of the different models of man and the world deployed at different times and in different societies, tells us also that it is the case – indeed, must be the case, in that Hegelian sense of ‘must’ which Berlin has so helpfully refused to dismiss – that not all values can be equally present to all outlooks. Moreover, intense consciousness of the plurality of values and of their conflict is itself a historical phenomenon, a feature of some ages (for instance, ours) rather than others. One thing, indeed, which can give us an insight into the point or claim of a certain value, its possible hold on our sentiments, is sympathetic understanding of a society which respected it with less pluralistic competition than it receives in ours.

The pluralism of values that Berlin advances is not just an application to ethics and political theory of the general anti-reductionist, anti-simplifying attitude in philosophy which he advances in the essay ‘Logical Translation’ (an essay which expresses very clearly some of the concerns of Oxford philosophy at that time). That general attitude appropriately gives way in the face of the demands of explanatory theory: it is obvious, indeed, that it has to give way in the face of theory, and the question in philosophy is how far explanation requires theory – a question to which present practice gives a much more positive answer than did the Oxford philosophy of the 50s. But the question in ethics, whether we should abandon the claims of some value which has force with us – abandon, for instance, considerations of loyalty or justice in the interests of general utility or benevolence – can hardly be a matter of explanatory theory. Philosophers have insisted, and still insist, that we encounter here the demands of another kind of theory, moral theory, which aims to systematise and simplify our moral opinions. But they rarely even try to answer a real question: what authority are theoretical tidiness or simplicity supposed to have against the force of concerns which one actually finds important? That question has no obvious answer, even after one has conceded considerable power (more, perhaps, than Berlin himself would concede) to philosophical theory in general.

It may be that there are no, or few, purely theoretical pressures to reduce the conflicts in our value-system. Berlin will say that there is a pressure to not reducing them, towards remaining conscious of these conflicts and not trying to eliminate them on more than a piecemeal basis: that pressure is the respect for truth. To deny the conflicts, indeed to try to resolve them systematically and once for all, would be to offend against something absolutely true about values. But then how are we to take the fact, already mentioned, that a high level of such conflict, and the consciousness of that, is a mark of some forms of life and some societies rather than others? Among the forms of life that support that kind of consciousness, a prominent position is needless to say occupied by the liberal society; and Berlin deploys the pluralism of values in defence of liberalism.

His defence of the liberal society is supported by the pluralism of values, I think, in more than one way. There is the obvious point that if there are many and competing genuine values, then the greater the extent to which a society tends to be single-valued, the more genuine values it neglects or suppresses. More, to this extent, must mean better. The point has strength even if we grant the important qualification that not all values can be pluralistically combined, and that some become very pale in too much pluralistic company. There are logical, psychological and sociological limits on what range of values an individual can seriously respect in one life, or one society respect in the lives of various of its citizens. (This is one thing that is being said by people who deny that liberal equality, for instance, is real equality – a point raised by the form that Berlin gives to equality in his discussion of it as one value among others.)

But there is a different kind of consideration, that the consciousness of the plurality of competing values is itself a good, as constituting knowledge of an absolute and fundamental truth. This is a good which, in the name of honesty, or truthfulness, or courage, may be urged against someone who recommends simplification of our values not, perhaps, as a theoretical necessity, but as a practical improvement. Here Berlin – in the last analysis, as thinkers of a rather different tendency put it – finds value in knowledge and true understanding themselves, and regards it as itself an argument for the liberal society that that society expresses more than any other does a true understanding of the pluralistic nature of values.

But what is that true understanding? What truth is it that is known to someone who recognises the ultimate plurality of values? In philosophical abstraction, it will be that there are such values, and, put in that blank way, it can be taken to speak for an objective order of values which some forms of consciousness (notably the liberal form) are better than others at recognising. But that way of putting it is very blank indeed. It is more characteristic of Berlin’s outlook, and more illuminating in itself, to say that one who properly recognises the plurality of values is one who understands the deep and creative role that these various values can play in human life. In that perspective, the correctness of the liberal consciousness is better expressed, not so much in terms of truth – that it recognises the values which indeed there are – but in terms of truthfulness. It is prepared to try to build a life round the recognition that these different values do each have a real and intelligible human significance, and are not just errors, misdirections or poor expressions of human nature. To try to build life in any other way would now be an evasion, of something which by now we understand to be true. What we understand is a truth about human nature as it has been revealed – revealed in the only way in which it could be revealed, historically. The truthfulness that is required is a truthfulness to that historical experience of human nature.

We can see, then, that in Berlin’s central conception of values and, connectedly, of humanity, there is an implicit appeal, once more, to historical understanding. We can perhaps see, too, how the development of his thought from general theory of knowledge to the history of ideas and the philosophy of history was not merely a change of interest; and that his complex sense of history is as deeply involved in his philosophy, even in its more abstract applications, as it is, very evidently, in his other writings, and in his life.

BERNARD WILLIAMS

¹ See Berlin’s own account, ‘Austin and the Early Beginnings of Oxford Philosophy’, in Sir Isaiah Berlin and others, Essays on J. L. Austin (Oxford, 1973), included in a later volume of the present selection, Personal Impressions (see p. vii above, note 1), and also Ayer’s autobiography, Part of my Life (London, 1977), p. 160.

The Purpose of Philosophy

WHAT IS THE subject-matter of philosophy? There is no universally accepted answer to this question. Opinions differ, from those who regard it as contemplation of all time and all existence – the queen of the sciences – the keystone of the entire arch of human knowledge – to those who wish to dismiss it as a pseudo-science exploiting verbal confusions, a symptom of intellectual immaturity, due to be consigned together with theology and other speculative disciplines to the museum of curious antiquities, as astrology and alchemy have long ago been relegated by the victorious march of the natural sciences.

Perhaps the best way of approaching this topic is to ask, what constitutes the field of other disciplines? How do we demarcate the province of, say, chemistry or history or anthropology? Here it seems clear that subjects or fields of study are determined by the kind of questions to which they have been invented to provide the answers. The questions themselves are intelligible if, and only if, we know where to look for the answers.

If you ask someone an ordinary question, say ‘Where is my coat?’, ‘Why was Mr Kennedy elected President of the United States?’, ‘What is the Soviet system of criminal law?’, he would normally know how to set about finding an answer. We may not know the answers ourselves, but we know that in the case of the question about the coat, the proper procedure is to look on the chair, in the cupboard, etc. In the case of Mr Kennedy’s election or the Soviet system of law we consult writings or specialists for the kind of empirical evidence which leads to the relevant conclusions and renders them, if not certain, at any rate probable.

In other words, we know where to look for the answer: we know what makes some answers plausible and others not. What makes this type of question intelligible in the first place is that we think that the answer can be discovered by empirical means, that is, by orderly observation or experiment, or methods compounded of these, namely those of common sense or the natural sciences. There is another class of questions where we are no less clear about the proper route by which the answers are to be sought, namely the formal disciplines: mathematics, for example, or logic, or grammar, or chess or heraldry, defined in terms of certain fixed axioms and certain rules of deduction etc., where the answer to problems is to be found by applying these rules in the manner prescribed as correct.

We do not know the correct proof of Fermat’s Theorem, for example, – no one is known to have found it – but we know along what lines to proceed: we know what kind of methods will, and what kind of methods will not, be relevant to the answer. If anyone thinks that answers to mathematical problems can be obtained by looking at green fields or the behaviour of bees, or that answers to empirical problems can be obtained by pure calculation without any factual content at all, we would today think them mistaken to the point of insanity. Each of these major types of questions – the factual and the formal – possesses its own specialised techniques: discoveries by men of genius in these fields, once they are established, can be used by men of no genius at all in a semi-mechanical manner in order to obtain correct results.

The hallmark of these provinces of human thought is that once the question is put we know in which direction to proceed to try to obtain the answer. The history of systematic human thought is largely a sustained effort to formulate all the questions that occur to mankind in such a way that the answers to them will fall into one or other of two great baskets: the empirical, i.e. questions whose answers depend, in the end, on the data of observation; and the formal, i.e. questions whose answers depend on pure calculation, untrammelled by factual knowledge. This dichotomy is a drastically over-simple formulation: empirical and formal elements are not so easily disentangled: but it contains enough truth not to be seriously misleading. The distinction between these two great sources of human knowledge has been recognised since the first beginning of self-conscious thinking.

Yet there are certain questions that do not easily fit into this simple classification. ‘What is an okapi?’ is answered easily enough by an act of empirical observation. Similarly ‘What is the cube root of 729?’ is settled by a piece of calculation in accordance with accepted rules. But if I ask ‘What is time?’, ‘What is a number?’, ‘What is the purpose of human life on earth?’, ‘How can I know past facts that are no longer there – no longer where?’, ‘Are all men truly brothers?’, how do I set about looking for the answer? If I ask ‘Where is my coat?’ a possible answer (whether correct or not) would be ‘In the cupboard’, and we would all know where to look. But if a child asked me ‘Where is the image in the mirror?’ it would be little use to invite it to look inside the mirror, which it would find to consist of solid glass; or on the surface of the mirror, for the image is certainly not on its surface in the sense in which a postage stamp stuck on it might be; or behind the mirror (which is where the image looks as if it were), for if you look behind the mirror you will find no image there – and so on.

Many who think long enough, and intensely enough, about such questions as ‘What is time?’ or ‘Can time stand still?’, ‘When I see double, what is there two of?’, ‘How do I know that other human beings (or material objects) are not mere figments of my own mind?’, get into a state of hopeless frustration. ‘What is the meaning of “the future tense”?’ can be answered by grammarians by mechanically applying formal rules; but if I ask ‘What is the meaning of “the future”?’ where are we to look for the answer?

There seems to be something queer about all these questions – as wide apart as those about double vision, or number, or the brotherhood of men, or purposes of life; they differ from the questions in the other basket in that the question itself does not seem to contain a pointer to the way in which the answer to it is to be found. The other, more ordinary, questions contain precisely such pointers – built-in techniques for finding the answers to them. The questions about time, the existence of others and so on reduce the questioner to perplexity, and annoy practical people precisely because they do not seem to lead to clear answers or useful knowledge of any kind.

This shows that between the two original baskets, the empirical and the formal, there is at least one intermediate basket, in which all those questions live which cannot easily be fitted into the other two. These questions are of the most diverse nature; some appear to be questions of fact, others of value; some are questions about words and a few symbols; others are about methods pursued by those who use them: scientists, artists, critics, common men in the ordinary affairs of life; still others are about the relations between various provinces of knowledge; some deal with the presuppositions of thinking, some with the nature and ends of moral or social or political action.

The only common characteristic which all these questions appear to have is that they cannot be answered either by observation or calculation, either by inductive methods or deductive; and, as a crucial corollary of this, that those who ask them are faced with a perplexity from the very beginning – they do not know where to look for the answers; there are no dictionaries, encyclopedias, compendia of knowledge, no experts, no orthodoxies, which can be referred to with confidence as possessing unquestionable authority or knowledge in these matters. Moreover some of these questions are distinguished by being general and by dealing with matters of principle; and others, while not themselves general, very readily raise or lead to questions of principle.

Such questions tend to be called philosophical. Ordinary men regard them with contempt, or awe, or suspicion, according to their temperaments. For this reason, if for no other, there is a natural tendency to try to reformulate these questions in such a way that all or at any rate parts of them can be answered either by empirical or formal statements; that is to say efforts, sometimes very desperate ones, are made to fit them into either the empirical or the formal basket, where agreed methods, elaborated over the centuries, yield dependable results whose truth can be tested by accepted means.

The history of human knowledge is, to a large degree, a sustained attempt to shuffle all questions into one of the two ‘viable’ categories; for as soon as a puzzling, ‘queer’ question can be translated into one that can be treated by an empirical or a formal discipline, it ceases to be philosophical and becomes part of a recognised science.¹ Thus it was no mistake to regard astronomy in, say, the early Middle Ages as a ‘philosophical’ discipline: so long as answers to questions about stars and planets were not determined by observation or experiment and calculation, but were dominated by such non-empirical notions as those, e.g., of perfect bodies determined to pursue circular paths by their goals or inner essences with which they were endowed by God or Nature, even if this was rendered improbable by empirical observation, it was not clear how astronomical questions could be settled: i.e. what part was to be played by observing actual heavenly bodies, and what part by theological or metaphysical assertions which were not capable of being tested either by empirical or by formal means.

Only when questions in astronomy were formulated in such a manner that clear answers could be discovered by using and depending on the methods of observation and experiment, and these in their turn could be connected in a systematic structure the coherence of which could be tested by purely logical or mathematical means, was the modern science of astronomy created, leaving behind it a cloud of obscure metaphysical notions unconnected with empirical tests and consequently no longer relevant to the new science, and so gradually relegated and forgotten.

So, too, in our own time, such disciplines as economics, psychology, semantics, logic itself, are gradually shaking themselves free from everything that is neither dependent on observation nor formal; if and when they have successfully completed this process they will be finally launched on independent careers of their own as natural or formal sciences, with a rich philosophical past, but an empirical and/or formal present and future. The history of thought is thus a long series of parricides, in which new disciplines seek to achieve their freedom by killing off the parent subjects and eradicating from within themselves whatever traces still linger within them of ‘philosophical’ problems, i.e. the kind of questions that do not carry within their own structure clear indications of the techniques of their own solution.

That, at any rate, is the ideal of such sciences; in so far as some of their problems (e.g. in modern cosmology) are not formulated in purely empirical or mathematical terms, their field necessarily overlaps with that of philosophy. Indeed, it would be rash to say of any developed high-level science that it has finally eradicated its philosophical problems. In physics, for instance, fundamental questions exist at the present time which in many ways seem philosophical – questions that concern the very framework of concepts in terms of which hypotheses are to be formed and observations interpreted. How are wave-models and particle-models related to one another? Is indeterminacy an ultimate feature of sub-atomic theory? Such questions are of a philosophical type; in particular, no deductive or observational programme leads at all directly to their solution. On the other hand, it is of course true that those who try to answer such questions need to be trained and gifted in physics, and that any answers to those questions would constitute advances in the science of physics itself. Although, with the progressive separation of the positive sciences, no philosophers’ questions are physical, some physicists’ questions are still philosophical.

This is one reason, but only one, why the scope and content of philosophy does not seem greatly diminished by this process of attrition. For no matter how many questions can be so transformed as to be capable of empirical or formal treatment, the number of questions that seem incapable of being so treated does not appear to grow less. This fact would have distressed the philosophers of the Enlightenment, who were convinced that all genuine questions could be solved by the methods that had achieved so magnificent a triumph in the hands of the natural scientists of the seventeenth and early eighteenth centuries.

It is true that even in that clear day men still appeared no nearer to the solution of such central, indubitably philosophical, because apparently unanswerable, questions as whether men and things had been created to fulfil a purpose by God or by nature, and if so what purpose; whether men were free to choose between alternatives, or on the contrary were rigorously determined by the causal laws that governed inanimate nature; whether ethical and aesthetic truths were universal and objective or relative and subjective; whether men were only bundles of flesh and blood and bone and nervous tissue, or the earthly habitations of immortal souls; whether human history had a discernible pattern, or was a repetitive causal sequence or a succession of casual and unintelligible accidents. These ancient questions tormented them as they had their ancestors in Greece and Rome and Palestine and the medieval west.

Physics and chemistry did not tell one why some men were obliged to obey other men and under what circumstances, and what was the nature of such obligations; what was good and what was evil; whether happiness and knowledge, justice and mercy, liberty and equality, efficiency and individual independence, were equally valid goals of human action, and if so, whether they were compatible with one another, and if not, which of them were to be chosen, and what were valid criteria for such choices, and how we could be certain about their validity, and what was meant by the notion of validity itself; and many more questions of this type.

Yet – so a good many eighteenth-century philosophers argued – a similar state of chaos and doubt had once prevailed in the realm of the natural sciences too; yet there human genius had finally prevailed and created order.

clarification