'The nineteenth-century fetishism of facts was completed and justified by a fetishism of documents. The documents were the Ark of the Covenant in the temple of facts. The reverent historian approached them with bowed head and spoke of them in awed tones. If you find it in the documents, it is so. But what, when we get down to it, do these documents - the decrees, the treaties, the rent-rolls, the blue books, the official correspondence, the private letters and diaries - tell us? No document can tell us more than what the author of the document thought what he thought had happened, what he thought ought to happen or would happen, or perhaps only what he wanted others to think he thought, or even only what he himself thought he thought. None of this means anything until the historian has got to work on it and deciphered it. The facts, whether found in documents or not, have still to be processed by the historian before he can make any use of them: the use he makes of them is, if I may put it that way, the processing process.

Let me illustrate what I am trying to say by an example which I happen to know well. When Gustav Stresemann, the Foreign Minister of the Weimar Republic, died in 1929, he left behind him an enormous mass - 300 boxes full - of papers, official, semi-official, and private, nearly all relating to the six years of his tenure of office as Foreign Minister. His friends and relatives naturally thought that a monument should be raised to the memory of so great a man. His faithful secretary Bernhard got to work; and within three years there appeared three massive volumes, of some 600 pages each, of selected documents from the 300 boxes… in the ordinary way the documents themselves would have mouldered away in some cellar or attic and disappeared for ever; or perhaps in a hundred years or so some curious scholar would have come upon them and set out to compare them with Bernhard's text. What happened was far more dramatic. In 1945,the documents fell into the hands of the British and American Governments, who photographed the lot and put the photostats àt the disposal of scholars at the Public Record Office in London and the National Archives in Washington, so that, if we have sufficient patience and curiosity, we can discover exactly what Bernhard did. What he did was neither very unusual nor very shocking. When Stresemann died, his western policy seemed to have been crowned with a series of brilliant successes - Locarno, the admission of Germany to the League of Nations, the Dawes and Young plans and the American loans, the withdrawal of allied occupation armies from the Rhineland. This seemed the important and rewarding part of Stresemann's foreign policy; and it was not unnatural that it should have been over-represented in Bernhard's selection of documents. Stresemann's eastern policy, on the other hand, his relations with the Soviet Union, seemed to have led nowhere in particular; and, since masses of documents about negotiations which yielded only trivial results were not very interesting and added nothing to Stresemann's reputation, the process of selection could be more rigorous. Stresemann in fact devoted a far more constant and anxious attention to relations with the Soviet Union, and they played a far larger part in his foreign policy as a whole, than the reader of the Bernhard selection would surmise. But the Bernhard volumes compare favorably, I suspect, with many published collections of documents on which the ordinary historian implicitly relies.

This is not the end of my story. Shortly after the publication of Bernhard's volumes, Hitler came into power. Stresemann's name was consigned to oblivion in Germany, and the volumes disappeared from circulation: many, perhaps most, of the copies must have been destroyed. …But in the west Stresemann's reputation stood high. in 1935 an English publisher brought out an abbreviated translation of Bernhard's work - a selection from Bernhard's selection; perhaps one-third of the original was omitted. Sutton, a well-known translator from the German, did his job competently and well. The Eng1ish version, he explained in the preface, was 'slightly condensed, but only by the omission of a certain amount of what, it was felt, was more ephemeral matter ... of little interest to English readers or students'. This again is natural 'enough. But the result is that Stresemann's eastern policy, already under-represented in Bernhard, recedes still further from view, and the Soviet Union appears in Sutton's volumes merely as an occasional and rather unwelcome intruder in Stresemann's predominantly western foreign policy. Yet it is safe to say that, for all except a few specialists, Sutton and not Bernhard - and still less the documents themselves – represents for the western world the authentic voice of Stresemann. Had the documents perished in 1945 in the bombing, and had the remaining Bernhard volumes disappeared, the authenticity and authority of Sutton would never have been questioned. Many printed collections of documents, gratefully accepted by historians in default of the originals, rest on no securer basis than this.

But I want to carry the story one step further. Let us forget about Bernhard and Sutton, and be thankful that we can, if we choose, consult the authentic papers of a leading participant in some important events of recent European history. What do the papers tell us? Among other things they contain records of some hundreds of Stresemann's conversations with the Soviet Ambassador in Berlin and of a score or so with Chicherin. These records have one feature in common. They depict Stresemann as having the lion's share of the conversations and reveal his arguments as invariably well put and cogent, while those of his partner are for the most part scanty, confused and unconvincing. This is a familiar characteristic of all records of diplomatic conversations. The documents do not tell us what happened, but only what Stresemann thought had happened, or what he wanted others to think, or perhaps what he wanted himself to think, had happened. It was not Sutton or Bernhard, but Stresemann himself, who started the process of selection. And if we had, say, Chicherin's records of these same conversations, we should still learn from them only what Chicherin thought, and what really happened would still have to be reconstructed in the mind of the historian. Of course, facts and documents are essential to the historian. But do not make a fetish of them. They do not by themselves constitute history; they provide in themselves no ready-made answer to this tiresome question 'What is history?'