I feel greatly privileged to have been invited to deliver a lecture devoted to the memory of a fellow-historian, who was so dedicated to her work and so humane a person as Dr Papiya Ghosh, and who has been so brutally snatched away from us. I deeply regret that owing to my age and health, it has not been possible for me to deliver this lecture in person, for which I seek your very kind indulgence.

When discussing a suitable theme for this lecture with Professor Bhuvan Chandel, I suggested, with an eye to the audience being mainly a gathering of philosophers, that I should speak on the Philosophy of History. I should hasten immediately to confess that I am not going to discuss how philosophers (such as Hegel and Marx, the latter in his capacity as a philosopher) have looked at History, nor even how the History of the past may be reshaped in the light of modern developments of philosophy. Even Post-modernism will be rather tangential to what I shall be speaking about.

In effect, what I propose doing is to start from the end of History, not from that of Philosophy, and, examining its raison d’être, go on to discuss how it is constructed, first by a collection of facts, which constitutes the research part, and then, by a selection and evaluation of those facts, which constitutes what is called interpretation. It is mainly in the latter sphere that History directly interacts with the domain of Philosophy. But here let me not dilate further on what I am going to say, and just proceed with my task.

First, let me begin with the question, why History? Why should we have it at all? Marc Bloch starts his last book, The Historian's Craft, by posing this precise question: of what use is history? My favourite answer – perhaps not original – is: why then does an individual have to have a memory? He may be very unimportant, his presence may not change circumstances at all, but for him his memory is of the very essence. It could be a memory of near-accidents, of meeting people, of what happened when he did something. Without such memory an individual cannot function, let alone avoid pitfalls. I suppose in the same way, groups of people, classes, or communities, and, in the modern world, peoples of nations and countries, need to make use of memory, or rather, make use of history as a source of memory. The more accurate our memory, i.e. our grasp of history, the more beneficial it will be for us as we view and draw lessons from our past errors, successes, and the reasons for them.

An individual has a memory of his own past conduct, and he has a memory of other people's past doings. But, just as one cannot be certain about the accuracy of a person's memory of his own action and the actions of others, one cannot be certain about the collective memories of groups of people of what happened to them or to others in the past. It is, therefore, as important for a whole people to make their memory correspond, as far as possible, to actual fact, just as it is necessary for an individual not to invent facts or let his memory be distorted by various complexes.

What people think is their history matters very much indeed, because it affects their practice. What they believe to be their past may not necessarily be true – like an individual's false memory, people can have a false history too. And that can exercise dangerous influences on their conduct. The creation of mythology may give us lessons for moral (and immoral!) conduct, but much else besides which can in real life generate divisive identities and mutual hostility among different sets of people, and one should not think that mythology can ever replace history to anyone's benefit. Just as false history cannot replace accurate history without great damage to a nation's conduct: of this Nazi Germany offers an unforgettable example. This explains why many Indian historians were so greatly perturbed during the BJP regime in our country, when a senseless glorification of India's ancient past was attempted.

Now, History like memory has one basic component, a large assemblage of facts. Here the most important question is the one I have just raised: the question of accuracy. But there is a second element equally essential; the arrangement and selection of facts, essential to the business of interpretation. Historical method embraces both an attempt at accuracy and at interpretation (or generalization, if you like). Interpretation involves not only the selection of facts but also the allotting of particular weightage to particular facts.

The problem with authors like Edward Said, it seems to me, is that they confuse the two aspects and also ignore their inter-relationships. Factual accuracy can be achieved by a continuous development of text criticism, archaeological discoveries, decipherment techniques, and so on. This constitutes the kernel of the scientific method, which, having been developed in modern Europe, has by now diffused all over the world. In my view Edward Said confuses this, namely the genesis of scientific method in the West, with assumptions of western supremacy.

He deals with the contributions made by Orientalism (his name for western scholarship on Asia and North Africa during the last three centuries) to the discovery of facts – a crucial contribution by them – in just one paragraph. But did not the discovery of old civilizations, languages, and texts altogether change existing interpretations? Indeed, Orientalists created by their discoveries the real groundwork for critiques of the reigning assumptions of western superiority though such assumptions might have been at the back of the mind of many of the Orientalists themselves. H.A.L. Fisher, a well-known British historian, had said that Europe discovered practically everything worth discovering, and not Asia or Africa. Orientalists found that Europe of itself did not discover many crucially important things, as Joseph Needham – surely an Orientalist by any definition one may adopt – showed so brilliantly in his Science and Civilization in China volumes. So actually, unless one developed and applied the historical method as the Orientalists in fact did, one cannot build any rational historical interpretation, or even oppose the assumptions of Western cultural hegemony that Edward Said protests against.

Once new facts are established, they may lead, as in the case we have just been discussing, to a change in the existing interpretations. With interpretation there certainly comes the issue of bias in the sphere of selection and generalization. Incidentally, a historian who copies others’ conclusions simply adopts their biases, and cannot claim to be unbiased just because he does not let his own bias intrude into his writing.

On the other hand, suppose we are dealing with a historian who is directly investigating a field. If he deals with a period in which much source material is available, then of course he will have to weigh the importance of each fact to judge whether it should enter his selection. There may be, in some cases, on the other hand, very little material on a particular aspect, but that aspect may be very important. So it is not just the quantity of material that may determine the weightage given to facts, but an understanding also of how historical processes take place, and which of them should be regarded as the more important ones. Such an understanding would depend on the historian's own personal views; and possibly, if he is trying to address an audience, he might have the motivation of tailoring his interpretation to what the audience wants to hear, or to what is more likely to appeal to his audience.

It is thus an undoubtedly complex matter we touch on, when we think of how biases of a historian develop. Nowadays, with the huge book market, historians are increasingly thinking of their audiences and how they can produce a bestseller or something that would appeal to the employment-controlling authorities. It may not happen consciously, but one must also understand that the market today (including the job market) is exercising a pull and shaping historians’ biases.

Marxist historians have a framework which could be flexible, and within which different weights could be assigned to different facts, although they would regard particular aspects as more important than, let us say, the mainstream western historians would do. Thus ideology also has an influence on one's selection of facts. In other words, there is the personal predilection of the historian, the views of the audience he is addressing, and the reigning ideological frameworks that may shape the bias of the historian. This certainly affects generalization, so that the same body of fairly accurate individual facts may lead quite validly to different interpretations, because of the different weights we assign to different parts of that evidence.

I would argue, however, that a point is reached when as our total knowledge of facts (or of what E.H. Carr called ‘historical facts’) grows, some earlier generalizations can no longer be sustained, as for example, the race-superiority theories. They have all been thrown out, because of our growing knowledge not only of different civilizations but also of genetics and even linguistics.

Now, in a sense when we touch on bias, we come very near to the area of a historian's personal philosophy. His own outlook reacts on his work, just as facts once discovered alter one's biases.

Let us look at why a historian chooses a subject. To give a very mundane example, because I had a Communist background, I chose to work on the agrarian system of Mughal India, at the same time that my friend, the late M. Athar Ali, chose the structure of nobility under Aurangzeb, because, being of a liberal persuasion, he wished to examine how far religious identities impinged on Mughal administrative functioning. We would not have chosen these different topics if our personal predilections were identical. So, even in research work, the very fact that one chooses a particular topic may reflect some previous presumptions about what is more significant in history. The fact that a historian like Emmanuel Le Roy Ladurie writes on the peasants of Languedoc, for example, simply means that he thinks this theme is important and chooses it in preference, say, to the biography of an aristocrat. Le Roy Ladurie's detailed research is not within a Marxist framework, but he certainly thinks that agrarian society is important. For Braudel it was the Mediterranean region that he took for his first great work, just because he was interested in multi-cultural regional civilizations. So let us admit that different historians with different ideological predilections even choose entirely different aspects for study, and thus, by this choice, declare, as it were, their bias. But still if one is confining oneself to accurate facts, there are limits beyond which generalizations cannot go. And I think there lies the difference between scientific or accurate history writing and non-historical approaches to the past.

I should enter a word here on the word ‘accuracy’. Historical method consists of techniques of ensuring that we understand earlier narratives better, identify by critical comparisons their biases, exaggerations, or omissions, in order to establish events as they in reality happened, or circumstances as they in reality shaped themselves. If historians cannot claim the exactitudes of laboratory experiments, where all conditions are controlled by the researcher, they can still claim reasonable approximations for their descriptions. Once this stage is reached, different interpretations would still be possible, but the range of such interpretations could be restricted and there is by no means any open sanction to say anything one likes.

My favourite example here is the late Professor R.C. Majumdar, a historian of high stature, with whom one may yet respectfully differ. It is not that his facts are wrong, but it is in the weightage he gives to certain facts on which one may basically differ with him. He was different from the RSS^{Footnote 2} people who produced the NCERT (National Council of Educational Research and Training) school textbooks under the BJP regime, as, unlike them, he tried to work with accurate facts (though possibly a biased selection of them), whereas the text-writers in the BJP regime just disregarded factual accuracy. There is a very vital difference here.

So certainly, as historical techniques improve, certain views or claims of historians or sections of historians are ruled out. What this means is that, given one's bias, one cannot say anything one likes and call it history. Indeed, the more accurate a historian's facts, the more limited would be the range of generalization available to him or her.

Let me here take up another way in which History's realm is broadened just because one has a particular bias. As the struggle for gender equality has grown, there is a natural urge to see how ordinary women have fared in history – their special trials and tribulations and also achievements. This was not a matter even liberal or Marxist historians were much concerned with when I began my research in the 1950s. In my book, the Agrarian System of Mughal India, published in 1963, women hardly appear, except for one or two statements referring to women holding zamindaris or landed properties. But it paid no attention to women as a sector in agrarian life.

In the last twenty or thirty years in Indian historiography there has been some change of approach in respect of women in history. I think there are two views about it. One is to confine the whole question of women to their status, and to ideas on how women were seen and depicted in texts. The other is to go beyond it, and to see how much was given to a woman to eat, to enquire about her livelihood, health, mortality rate, and so on. In other words, the actual position of women in different classes, because of course in the past women did not often see themselves as of one class or category, as they may today tend to do. The aristocratic woman, however secluded, would have nothing to do with the slave girl. And what was the position of the slave girl? Her position was practically always worse than that of the man slave. This struck me when I was reading the translations of Buddhist Tipitaka texts: in these the slave woman is regarded as the unhappiest person on earth. The fact did not strike me at first that these texts were not referring to men slaves, but to the woman slave alone. The woman slave had practically no rights; in Muslim law too she has no sexual protection. It would be true to say that such an understanding of women's travails is not to be found in most works of history written in India four decades ago or earlier.

Should women find a place in general historical writing, or be written about as a separate category? My plea would be that both should be done. Since many general things about technology, economic life, and society are not very clear, it is first important to know the details about women and then relate their condition to the general social framework. We see that in European history too, there are separate researches carried out on women of a particular region, class, town, or countryside. In any case, one cannot have a history of women without understanding how men were treating them, and how men also treated each other.

I think I have said enough now about the interaction between discovery of historical facts and the development of new ideas. All ideas are affected by how historical facts are seen; but there are many other sources too of change in ideas, such as economic and political pressures, struggles among social classes, new social classes, new social aspirations, altering religious beliefs, and, not the least, the ideas and insights of individual thinkers. These, in turn, affect what areas of history we research in more than other areas, and so shape the way we reconstruct History.

In modern times we seek increasingly to gather scattered ideas and shape them into systematic frameworks. Marxism is one philosophy where a comprehensive systematization is attempted. Marxism has been an attractive philosophy to many because it emphasizes what happens to the common people, how they were and are exploited, who exploited and exploits them, and how conditions are growing in which they can liberate themselves. This naturally assigns greater importance to economic history, but I should hasten to disavow the belief that Marxism implies any belief in economic determinism.

When Marx speaks of the material basis and superstructure, he necessarily includes mind (ideas) in the material basis as well. When one is speaking of production, it is impossible to speak of it being carried out without human skill, technology, and science. So the whole debate about materialism, as if matter alone is important, is misleading, and misrepresents Marx's position. He uses ‘materialism’ because that is the immediate word at hand, but he doesn’t mean that the material conditions exclude human skill, labouring power, science and technology – which are all products of the human mind or are controlled by it.

It is obviously very important to pay attention to ideas and their role; without that one cannot understand the rise of capitalism. For its emergence and growth, capitalism was dependent on capital accumulation, colonial conquests, the expropriation of the peasantry. All that is true, but it was also dependent on technology, and that was given to it by the scientific revolution. For all these reasons, the ideological battleground must always remain important in History.

In western historiography, the Marxist approach was getting greater and greater attention in the 1960s and 1970s. There were several reasons for this. One was that economic history was becoming the centrepiece of new historical research. The many debates that arose out of detailed historical work and quantitative work in particular were essentially about economic history. Political history cannot be quantitatively analysed, but economic trends can. Many questions that Marxists asked could be tested quantitatively. Of course there were debates in other disciplines like sociology also. In post-World War II sociology, Marx began to be considered as an important sociological figure. This extended even to psychology (re: Marx's ideas on ‘alienation’).

Secondly, there was I think a greater concern among historians and economists both from the welfare side and the socialist side that things which are important from the perspective of the poor, needed to be studied.

It seems to me that there has subsequently been a move away from this position. The first push came, I think, from a tendency within western historiography to defend the past of Europe in a world increasingly critical of it. This can be seen in the development of certain trends in British historical writing on India. We will not go into the motives in detail, but even if one were a liberal English historian in the so-called post-colonial era, one would not like to have a dark picture painted of British-ruled India, and would have liked to see if it couldn’t be painted differently. In England, where Indian studies suffered a great decline in the 1950s and 1960s, there came about a renewed interest in India as it was under the Raj. Such scholars got a particular engine and method for their cause in the ‘structural analysis’ approach advocated by Sir Lewis Namier. Actually, the approach was misnamed, because what Namier and his followers dealt with was the study of private papers and individual motivation. If you could ensure that Indian nationalist leaders had motives of personal careers, or caste and community affiliations, then a different picture emerged of Indian nationalism; and British rule which these motivated leaders criticized would not look so bad after all. These historians never or seldom handled the question of tribute (‘drain of wealth’ from India to Britain) or India's ‘de-industrialization’ in the nineteenth century: their concern was mostly with individuals. Thus, the national movement appeared to be based not on real popular grievances, but on elite-manufactured ones. This is roughly the position of the Cambridge school, represented by Gallagher and Anil Seal, C. Bayly, Judith Brown, and others.

A decade or two behind the Cambridge School came the Subalterns. The sources of the Subaltern trend in historiography were different. As far as its founder, Ranajit Guha, is concerned, his first book, A Rule of Property for Bengal, arose out of dissatisfaction with the nationalist historian N.K. Sinha. Sinha held that the Permanent Settlement in Bengal was instituted by the British to ensure the maximization of land revenue collection, but because of various factors, like movements of prices, it did not work so well for the British. Ranajit Guha argued that Sinha ignored the ideological fact that colonialism wanted to create landed property, and his own book is not really on economic history, but is essentially a study in intellectual history, of how this idea of landed property implanted in Bengal. Guha later developed his earlier approach into a general hostility towards mainstream Indian historiography.

He went on to hypothesize for modern India a struggle among three elements – ‘colonial elites’, ‘Indian elites’, and ‘subalterns’. It should be noted that his definition of subaltern is completely different from that of the Italian communist thinker Gramsci, who said that subaltern classes actually help to reinforce the hegemony of the ruling classes. Guha's definition equated ‘subaltern’ with ‘subordinate classes’, though this was linguistically wrong, because the Oxford English Dictionary defines ‘subaltern’ as a ‘subordinate clause’ not ‘class’. In logic the main argument is supported by a subordinate clause; so in fact subalterns should form a subordinate support to the cause of the ruling classes. But for Guha, from subordinate classes they became the main resisting classes. And then he created a picture without any historical basis whatsoever, of the subalterns being communities, which are not economic classes but only those whose members do not have an English or modern education. So a big landlord who had not passed high school is subaltern because he is not influenced by ‘elite’ ideas! Indian elites are not capitalists, they can also be workers too if they are influenced by elite ideas. This collection of assertions set up initially as premises was built up into a theology under whose influences ‘Subaltern’ historians wrote papers, mainly apparently to use the word ‘subaltern’, and decry the role of the ‘nationalist-elite’ leadership in colonial India.

The great weakness with the Subaltern theorists is that they do not deal with aggregates. Therefore, the colonial tribute does not come under their scanner nor does de-industrialization; a tribal person would not obviously have directly seen either tribute or de-industrialization. Even the peasant could only have seen the impact of direct taxation. A landless labourer may not have seen it as he would not have known why his real wages were falling. So the total rejection of economic statistics of higher magnitudes results in the total ignoring of any question as to why and how India was exploited by colonialism. This self-imposed blindness is basic to Subaltern historiography. The Subaltern approach actually fits in very well with the Cambridge school because both attack the Indian ‘elites’ whose members worked out the ways Britain was exploiting India and created ‘Economic Nationalism’, on which Bipan Chandra has written a classic work. To the Sublaterns this seems to have just been an elite exercise.

The leading Subaltern scholar now seems to be Professor Partha Chatterjee. He and his colleagues even see communalism as a ‘subaltern’ phenomenon, and argue that the ‘Nehruvian Marxists’ are wrong in thinking that the Indian people are not communal. These are extremely disingenuous statements, though they have got considerable support abroad. Edward Said also wrote in praise of the ‘Subalterns’, but he did warn them that many of their ideas could be ‘complicit with neo-colonialism’!

The Subalterns have found post-Modernism very useful, because post-modernism also rejects statistical aggregates and any large economic frameworks. One of the early works anticipating post-modernism was by Louis Dumont. In his book Homo Hierarchicus he rejects the notion that India has a history, or that it has an economic history, or even that economic ideas can be applied to Indian society. All this has been completely absorbed by the Subalterns who are pleased to hitch on to Dumont's emphasis on the theological underpinning of caste. Thus, not only is communalism a subaltern phenomenon, but so are tribalism and casteism. Dumont also fits in with post-modernism, because the latter rejects the ‘meta narrative’. The rejection of the meta-narrative means that one cannot deal with large universally recognizable cultural or economic factors, one can only deal with individuals and individual communities, cultures, and regions. One certainly cannot apply a common, universally applicable scientific method. All this fits in with not only the Subaltern approach, but also the religious communal approach to History, so much in evidence in India. Edward Said was once a favourite author of the Indian right-wing RSS journalists, and they frequently cited him in their battle with Indian secularism and modern values.

I suppose I have said enough about the kinds of criticisms which, in my view, certain contemporary trends in Indian historiography legitimately invite. I would now venture to suggest that out of the continuous and ongoing relationship between historical facts and evolving social values (which form a domain shared by both philosophy and history), certain fundamental propositions in relation to History arise, which may tentatively be set forth as follows.

History, first of all, spans a range between the individual and the whole of humankind. The history of particular divisions into which humanity may be divided, such as nations, cultures, communities, tribes, localities, has to be undertaken without forgetting both the whole to which the divisions belong and the primary parts into which they in turn may dissolve. In practical terms it imposes on the historian the obligation of exercising a clinical degree of impartiality. It is not the historian's task to glorify a particular section of humanity. In the last forty years or so there has been the development of a trend in archaeology, originating from the writings of the US archaeologist Louis Binford, that was first called New Archaeology, but now is known as Processual Archaeology. One of its pillars is the assumption that every change in any locality or region comes about primarily as a result of independent response to environment rather than as a consequence of any import from any other territorial source or culture. So strong became the tirade against ‘diffusion’ that archaeologists of the older persuasion often tended to be dubbed ‘diffusionists’, and for that reason ignored. Such a theory denies in effect, that all segments of humanity have, down the ages, learnt from each other, a process only hastened and intensified as mutual contacts have increased. So strong became the prejudice generated against diffusion by New/Processual Archaeology that it tended to reinforce historical chauvinism in national and continental histories. It can be seen that it fits well with the tendency in Post-modernism to deny the possibility of universal values and to uphold cultural autonomies. Here, however, once again, facts, forming the hard evidence, are proving to be an effective censor. A major refutation of Processual Archaeology and anti-diffusionism from some of the major archaeologists all over the world came in the compilation The Origin and Spread of Agriculture and Pastoralism in Eurasia, edited by David R. Harris (1996). In history proper, the work of historians of inter-cultural intercourse and of technological transmissions, as seen in the great volumes written or sponsored by Joseph Needham on China and the studies of Lynn White Jr – to take just two examples out of a multitude – has established that in fact the greatness of a culture – its potentiality of growth, that is – is to be judged mainly by the degree of its openness to external influences. In other words, a historian has to seek dispassionately the origins of any new technique, any new discovery, any new belief or philosophical insight, wherever they may lie, and not become an advocate, as if in court, for a particular country, culture, or religion. Let me give two examples. There is today very little recognition among historians of the respective ‘semitic’ religions, Judaism, Christianity, and Islam, of the Zoarastrian influence that may have gone into shaping their notions of God, Satan, Angels, Prophet, Judgement Day, the narrow bridge to Paradise and Paradise itself. The other example comes from the trend among some Indian historians, to deny the external origins of the Indo-Aryan languages and to underplay the Mesopotamian and Greek influences on ancient Indian astronomy. In other words, even when, as must usually be the case, a historian concerns himself with a particular region or culture, his context must remain the whole of humanity, and he should not avoid tracing chains of diffusion to external sources.

A corollary of this must be the rejection of the argument, promoted by Edward Said's work on Orientalism, that the history of a culture or religion can only be studied ‘on its own terms’, and so preferably by a member or follower of that culture or religion. There cannot, however, be separate ‘terms’ laid down for the historian for the study of a particular area, faith, or community. This implies that, in so far as a historian is concerned, he should not allow his judgement to be affected by his or other people's faith. Alas, he does not have the freedom that the judges of one of our High Courts have just displayed in preferring faith over fact, in the Ayodhya case. One cannot stress too often that the canons of historical method, the principles of texts and narratives, the rationales of analysis, etc., are all universally applicable.

Universality of method does not, however, necessarily do away with the danger posed by insularity of approach. A historian's standpoint is often defined by his own environment, which may influence unconsciously how he looks at the world. There are few historians today of the stature that rightfully belongs to Eric Hobsbawm. Hobsbawm belongs to a country (England) where ahead of, and along with, the Industrial Revolution of the eighteenth century, the peasants were destroyed as a class by the process known as Enclosures and so it comes naturally to him to ask in his history of the twentieth century, how peasants would now fare in an industrializing world. Yet a reader of his book would not find in it any recognition of one great achievement of the twentieth century, namely, the destruction and severe constriction of landlordism over an area containing the majority of mankind, nor of the vast social and economic consequences that have resulted from this change. One feels that Marc Bloch, belonging to France, a country yet possessed of a large class of peasantry, could have more readily noted the phenomenon and given adequate importance to it. In other words, one must surely try ever more to understand different events and circumstances by putting oneself in one's mind in the place of people of other countries and, indeed, civilizations, to ensure that a common past is seen from as many diverse angles as is humanly possible.

From this rather large proposition of taking the whole of humanity as ever the context, while adopting a universally valid critical approach, I should like to raise what may be a more contentious point within what I am calling the Philosophy of History. The historical narrative must alter in its areas of emphasis and selection of details, according as universally recognized values change and develop. Once we recognize that everyone has a right to vote, we are bound to acknowledge that every one has a right to a share of history whether of his class, territory, or community, even within the history of his nation. However one may admire Gibbon's Decline and Fall of the Roman Empire, today's historian is bound to see history differently, for his concerns are far broader than those of Gibbon. I am reminded here of M.I. Rostovtzeff. A Russian exile and critic of the Soviet Revolution, he went to History to show how wrong the Marxist approach was. Yet his Social and Economic History of the Roman Empire (1926) proved to be a fundamental study of classes and class struggle within the Roman Empire – a Marxist text, almost – because to him history was wider than a narrative of Caesars, military campaigns, and Greek and Latin literatures. In the last fifty years or so there have been two movements among historians, one called ‘History from Below’, originating with the interest of British historians (notably R. Thompson) in working-class history, and the other called the ‘New History’ of the French Annales School, taking up the history of ‘marginal’ groups and classes.

Not the least part of this movement in historiography, which I have already touched upon, has been the attention beginning to be paid to the narrative of women as the repressed gender. What played the role of a catalyst in this was, perhaps, S. de Beauvoir's book, The Second Sex, whose English translation appeared in 1953. Even Marxist historians earlier had neglected the matter of women's history, assuming in effect that the interests of men and women in the same class were identical. What was overlooked was surely the fact that gender repression has been a major ideological factor, which seemed to earn legitimacy for any social order however unjust. Men of the lowest class could feel that there were still some persons (women) who were inferior to them, just as women of higher ranks (even those marginally higher) would feel superior to their lower-placed sisters. Gender inequality was thus not only an inherent part of general social inequality, but partly its ideological sustainer.

It is thus clear that the philosophy of History must embrace whatever flows from the recognition of equality as a basic principle (even in the diluted form propounded by John Rawls); and past events must be judged according to how they affected different ranks or classes of the people at that time, immediately or by distant consequence.

It follows that the historical narrative must change not only as more facts are discovered, but also as our ideas and values change. This may appear to some as a rather disturbing notion; but it is, perhaps, as inevitable as the fact that philosophers cannot for ever be satisfied to remain within domains presently reached, or that Physics and Chemistry stagnate at the point where their present discoveries have taken them. If whatever one writes today becomes irrelevant as humanity changes for the better tomorrow, that, I suppose, is not to be mourned, but welcomed. Had not Marx said: ‘Philosophers have hitherto interpreted the world; the point, however, is to change it.’ And history will change as surely as the philosophers and the world both keep changing.