The Proper Study of Mankind

by Isaiah Berlin

  To suppose, then, that there have been or could be ages without political philosophy is like supposing that, as there are ages of faith, so there are or could be ages of total disbelief. But this is an absurd notion: there is no human activity without some kind of general outlook. Scepticism, cynicism, refusal to dabble in abstract issues or to question values, hard-boiled opportunism, contempt for theorising, all the varieties of nihilism are, of course, themselves metaphysical and ethical positions, committal attitudes. Whatever else the existentialists have taught us, they have made this fact plain. The idea of a completely wertfrei theory (or model) of human action (as contrasted, say, with animal behaviour) rests on a naïve misconception of what objectivity or neutrality in the social studies must be.

  V

  The notion that a simile or model drawn from one sphere is necessarily misleading when applied to another, and that it is possible to think without such analogies in some direct fashion – ‘face to face’ with the facts – will not bear criticism. To think is to generalise, to generalise is to compare. To think of one phenomenon or cluster of phenomena is to think in terms of its resemblances to and differences from others. This is by now a hoary platitude. It follows that without parallels and analogies between one sphere and another of thought and action, whether conscious or not, the unity of our experience – our experience itself – would not be possible. All language and thought is, in this sense, necessarily ‘metaphorical’. The models, once they are made conscious and explicit, may turn out to be obsolete or misleading. Yet even the most discredited among these models in politics – the social contract, patriarchalism, the organic society and so forth – must have started with some initial validity to have had the influence on thought that they have had.

  No analogy powerful enough to govern the concepts of generations of men can have been wholly specious. When Jean Bodin or Herder or the Russian Slavophils or the German sociologist Tönnies transfer the notion of family nexus to political life, they remind us of aspects of relationships between men united by traditional bonds, or bound by common habits and loyalties, which had been misrepresented by the Stoics or Machiavelli or Bentham or Nietzsche or Herbert Spencer. So, too, assimilation of law to a command issued by some constituted authority in any one of the three types of social order distinguished by Max Weber throws some light on the concept of law. Similarly, the social contract is a model which to this day helps to explain something of what it is that men feel to be wrong when a politician pronounces an entire class of the population (say capitalists or blacks) to be outside the community – not entitled to the benefits conferred by the State and its laws. So too Lenin’s image of the factory which needs no supervision by coercive policemen after the State has withered away; Maistre’s image of the executioner and his victims as the corner-stone of all authority, or of life as a perpetual battlefield in which only terror of supernatural power keeps men from mutual extermination; the State’s role as traffic policeman and night-watchman (Lassalle’s contemptuous description of the liberal ideal); Locke’s analogy of government with trusteeship; the constant use by Burke and the entire romantic movement of metaphors drawn from organic growth and decay; the Soviet model of an army on the march, with its accompanying attributes and values, such as uncritical loyalty, faith in leadership, and military goals such as the need to overtake, destroy, conquer some specified enemy – all these illuminate some types of social experience.

  The great distortions, the errors and crimes that have sought their inspiration and justification in such images, are evidence of mechanical extrapolation, or over-enthusiastic application to the whole, of what, at most, explains a sector of life. It is a form of the ancient fallacy of the Ionian philosophers, who wanted a single answer to the question ‘What are all things made of?’ Everything is not made of water, nor fire, nor is explained by the irresistible march towards the world State or the classless society. The history of thought and culture is, as Hegel showed with great brilliance, a changing pattern of great liberating ideas which inevitably turn into suffocating strait-jackets, and so stimulate their own destruction by new, emancipating, and at the same time enslaving, conceptions. The first step to the understanding of men is the bringing to consciousness of the model or models that dominate and penetrate their thought and action. Like all attempts to make men aware of the categories in which they think, it is a difficult and sometimes painful activity, likely to produce deeply disquieting results. The second task is to analyse the model itself, and this commits the analyst to accepting or modifying or rejecting it, and, in the last case, to providing a more adequate one in its stead.

  It is seldom, moreover, that there is only one model that determines our thought; men (or cultures) obsessed by single models are rare, and while they may be more coherent at their strongest, they tend to collapse more violently when, in the end, their concepts are blown up by reality – experienced events, ‘inner’ or ‘outer’, that get in the way. Most men wander hither and thither, guided and, at times, hypnotised by more than one model, which they seldom trouble to make consistent, or even fragments of models which themselves form a part of some none too coherent or firm pattern or patterns. To drag them into the light makes it possible to explain them and sometimes to explain them away. The purpose of such analysis is to clarify; but clarification may expose shortcomings and subvert what it describes. That has often and quite justly been charged against political thought, which, at its best, does not disclaim this dangerous power. The ultimate test of the adequacy of the basic pattern by which we think and act is the only test that common sense or the sciences afford, namely, whether it fits in with the general lines on which we think and communicate; and if some among these in turn are called into question, then the final measure is, as it always must be, direct confrontation with the concrete data of observation and introspection which these concepts and categories and habits order and render intelligible. In this sense, political theory, like any other form of thought that deals with the real world, rests on empirical experience, though in what sense of ‘empirical’ still remains to be discussed.

  VI

  When one protests (as we ourselves did above) that the application of such (social or political) models or combinations of overlapping models, which at most hold a part of our experience, causes distortion when applied beyond it, how do we set about justifying this charge? How do we know that the result is distortion? We usually think this because the universal application of a simile or a pattern – say that of the general will, or the organic society, or basic structure and superstructure, or the liberating myth – seems to those who reject it to ignore something that they know directly of human nature and thereby to do violence to what we are, or what we know, by forcing it into the Procrustean bed of some rigid dogma; that is to say, we protest in the name of our own view of what men are, have been, could be.

  How do we know these things? How do we know what is and what is not an adequate programme for human beings in given historical circumstances? Is this knowledge sociological, or psychological? Is it empirical at all, or metaphysical and even theological? How do we argue with those whose notions are different from ours? Hume, Helvétius, Condorcet, Comte are clear that such knowledge must be based on empirical data and the methods of the natural sciences; all else is imaginary and worthless.

  The temptation to accept this simple solution was (and is) very great. The conflict of the rival explanations (or models) of social and individual life had, by the late eighteenth century, grown to be a scandal. If one examines what answers were offered, let us say, between the death of Newton and the birth of Darwin, to a central political question – why anyone should obey anyone else – the babel of voices is appalling, perhaps the most confused in recorded history. Some said that I should obey those rules or institutions submission to which alone would fulfil my nature, with the rider that my needs and the correct path to their satisfaction were clear only to those privileged observers who grasped at least some part
of the great hierarchy of being. Others said that I should obey this or that authority or law because only in that way could I (without aid of experts) fulfil my ‘true’ nature, or be able to fit into a harmonious whole. Some supposed this whole to be static; others taught that it was dynamic, but could not agree on whether it moved in recurrent cycles, or a straight, or spiral, or irregular evolutionary line, or by a series of oscillations leading to ‘dialectical’ explosions; or again, whether it was teleological or functional or causally determined.

  Some conceived the ultimate universal pattern in mechanistic, others in organic, others still in aesthetic terms. There were those who said that men must obey because they had promised to do so, or others had promised on their behalf; or that they were behaving as if they had promised and this was tantamount to having promised, whether they admitted this or no; or, if this seemed unconvincing, that it were best that they should behave as if they had so promised, since otherwise no one would know where he was and chaos would ensue. Some told men to obey because they would be happier if they did, or because the majority, or all men, would be happier; or because it was God’s will that they should obey, or the will of the sovereign, or of the majority, or of the best or wisest, or of history, or of their State, or their race, or their culture, or their Church.

  They were told also that they must obey because the natural law laid down that they must do so, but there were differences about how the precepts of natural law were to be discovered, whether by rational or by empirical means, or by intuition, and again, by common men or only by the experts; the experts in their turn were identified by some with natural scientists, by others with specialists in metaphysics or theology, or perhaps in some other discipline – mass psychology, mystical revelation, the laws of history, of economics, of natural evolution, of a new synthesis of all or some of these. Some people supposed that truth in these matters could be discovered by a faculty which they called moral sense, or common sense, or the perception of the fitness of things, or that it consisted in what they had been told by their parents or nurses, or was to be found in accepted views which it was mere perversity to question, or came from one or other of many sources of this sort which Bentham mocks at so gaily and effectively. Some (and perhaps these have always been the majority) felt it to be in some degree subversive to raise such questions at all.

  This situation caused justified indignation in a country dominated by free enquiry and its greatest triumph, Newtonian science. Surely this monstrous muddle could be cleared away by the strong new broom of scientific method – a similar chaos had, after all, not so long ago prevailed in the natural sciences too. Galileo and Newton – and the light of reason and experiment – had silenced for ever the idle chatter of the ignoramus, the dark muttering of the metaphysician, the thunder of the preacher, the hysterical shrieks of the obscurantist. All genuine questions were questions of discoverable fact – calculemus was to be the motto of the new method;1 all problems must be so reformulated that inspection of the facts – aided by mathematical techniques – would answer them decisively, with a clear, universally valid, empirical statement of verifiable fact.

  VII

  Nevertheless, attempts by the philosophes of the eighteenth century to turn philosophy, and particularly moral and political philosophy, into an empirical science, into individual and social psychology, did not succeed. They failed over politics because our political notions are part of our conception of what it is to be human, and this is not solely a question of fact, as facts are conceived by the natural sciences; nor the product of conscious reflection upon the specific discoveries of anthropology or sociology or psychology, although all these are relevant and indeed indispensable to an adequate notion of the nature of man in general, or of particular groups of men in particular circumstances. Our conscious idea of man – of how men differ from other entities, of what is human and what is not human, or inhuman – involves the use of some among the basic categories in terms of which we perceive and order and interpret data. To analyse the concept of man is to recognise these categories for what they are. To do. this is to realise that they are categories, that is, that they are not themselves subjects for scientific hypotheses about the data which they order. The analogy with the sciences which dominates the pre-Kantian thinkers of the eighteenth century – Locke, Hume and Condillac, for example – is a typical misapplication of a model that works in one sphere to a region where it will obscure at least as much as it illuminates.

  Let me try to make this more specific. When the theological and metaphysical models of the Middle Ages were swept away by the sciences of the seventeenth and eighteenth centuries, they disappeared largely because they could not compete with new disciplines in describing, predicting, controlling the contents of the external world. To the extent to which man was regarded as an object in material nature the sciences of man – psychology, anthropology, economics, sociology and so on – began to supplant their theologico-metaphysical predecessors. The questions of the philosophers were affected by this: some were answered or rendered obsolete; but some remained unanswered. The new human sciences studied men’s actual habits; they promised, and in some cases provided, analyses of what men said, wanted, admired, abhorred; they were prepared to supply empirical evidence for this, or experimental demonstration; but their efforts to solve normative problems were less successful. They tried to reduce questions of value to questions of fact – of what caused what kind of men to feel or behave as they did in various circumstances. But when Kant or Herder or Dostoevsky or Marx duly rejected the Encyclopaedists’ answers, the charge against them was not solely that of faulty observation or invalid inference; it was that of a failure to recognise what it is to be a man, that is, failure to take into account the nature of the framework – the basic categories – in terms of which we think and act and assume others to think and act, if communication between us is to work.

  In other words, the problem the solutions of which were found insufficient is not in the usual sense empirical, and certainly not formal, but something that is not adequately described by either term. When Rousseau (whether he understood him correctly or not) rejected Hobbes’s account of political obligation on the ground that Hobbes seemed to him to explain it by mere fear of superior force, Rousseau claimed not that Hobbes had not seen certain relevant empirical, psychologically discoverable, facts, nor that he had argued incorrectly from what he had seen – but that his account was in conflict with what, in thinking of human beings as human, and distinguishing them, even the most degraded among them, not only in explicit thought, but in our feelings and in our action, from beings that we regard as inhuman or non-human, we all know men to be. His argument is not that the facts used to construct Hobbes’s model had gaps in them, but that the model was inadequate in principle; it was inadequate not because this or that psychological or sociological correlation had been missed out, but because it was based on a failure to understand what we mean by motive, purpose, value, personality and the like.

  When Kant breaks with the naturalistic tradition, or Marx rejects the political morality of Bentham, or Tolstoy expresses a low opinion of the doctrines of Marx, they are not complaining merely of empirical ignorance or poor logic or insufficient experimental evidence, or internal incoherence. They denounce their adversaries mainly for not understanding what men are and which relationships between them – or between them and outside forces – make them men; they complain of blindness not to the transient aspects of such relations, but to those constant characteristics (such as discrimination of right from good for Kant, or, for Marx, systematic self-transmutation by their own labour) that they regard as fundamental to the notion of man as such. Their criticisms relate to the adequacy of the categories in terms of which we discuss men’s ends or duties or interests, the permanent framework in terms of which, not about which, ordinary empirical disagreements can arise.

  What are these categories? How do we discover them? If not empirically, then by what means? How un
iversal and unchanging are they? How do they enter into and shape the models and paradigms in terms of which we think and respond? Do we discover what they are by attention to thought, or action, or unconscious processes, and how do we reconcile these various sources of knowledge? These are characteristically philosophical questions, since they are questions about the all but permanent ways in which we think, decide, perceive, judge, and not about the data of experience – the items themselves. The test of the adequate working of the methods, analogies, models which operate in discovering and classifying the behaviour of these empirical data (as natural science and common sense do) is ultimately empirical: it is the degree of their success in forming a coherent and enduring conceptual system.

  To apply these models and methods to the framework itself by means of which we perceive and think about them is a major fallacy, by the analysis of which Kant transformed philosophy. In politics it was committed (by Hume and Russell, for example) when enquiry into the empirical characteristics of men was confounded with the analysis of the notion of man (or ‘self’ or ‘observer’ or ‘moral agent’ or ‘individual’ or ‘soul’) in terms of which the empirical characteristics were themselves collected and described. Kant supposed these categories to be discoverable a priori. We need not accept this; this was an unwarranted conclusion from the valid perception that there exist central features of our experience that are invariant and omnipresent, or at least much less variable than the vast variety of its empirical characteristics, and for that reason deserve to be distinguished by the name of categories. This is evident enough in the case of the external world: the three-dimensionality of (psychological, common-sense) space, for example, or the solidity of things in it, or the ‘irreversibility’ of the time-order, are among the most familiar and inalienable kinds of characteristics in terms of which we think and act. Empirical sciences of these properties do not exist, not because they exhibit no regularities – on the contrary they are the very paradigm of the concept of regularity itself – but because they are presupposed in the very language in which we formulate empirical experience. That is why it seems absurd to ask for evidence for their existence, and imaginary examples are enough to exhibit their structure; for they are presupposed in our commonest acts of thought or decision; and where imaginary examples are, for the purpose of an enquiry, as good as, or even better than, empirical data drawn from actual experience, we may be sure that the enquiry is not, in the normal sense, an empirical one. Such permanent features are to be found in the moral and political and social worlds too: less stable and universal, perhaps, than in the physical one, but just as indispensable for any kind of intersubjective communication, and therefore for thought and action. An enquiry that proceeds by examples, and is therefore not scientific, but not formal, that is, deductive, either, is most likely to be philosophical.