[In the following excerpt, Woodman suggests that Tacitus is better read as a poet than a traditional historian.]

It was on the last Monday in January exactly fifty years ago that Adolf Hitler became Chancellor of Germany. The event has been described by Alan Bullock in his celebrated book:¹

During the morning a silent crowd filled the street between the Kaiserhof and the Chancellery. … At a window of the Kaiserhof, Röhm was keeping an anxious watch on the door from which Hitler must emerge. Shortly after noon a roar went up from the crowd: the Leader was coming. He ran down the steps to his car and in a couple of minutes was back in the Kaiserhof. As he entered the room his lieutenants crowded to greet him. The improbable had happened: Adolf Hitler, the petty official's son from Austria, the down-and-out of the Home for Men, the despatch-runner of the List regiment, had become Chancellor of the German Reich.

This account is not without its drama. There is a brief reference to the waiting crowd and its reaction to Hitler's appearance; and there is a poignant contrast, which Bullock underlines by the word ‘improbable’, drawn between Hitler's sudden eminence and three of the insignificant positions he had occupied earlier in his life. But these are the only emotional flashes which Bullock has permitted to obtrude into his account of that momentous day. The keynotes are stylistic reserve and brevity. The whole account amounts to only one paragraph of ten lines in the Pelican edition.

The events of that same day have also been described by William Shirer in his equally celebrated book on the Third Reich:²

Shortly before noon on Monday, January 30, 1933, Hitler drove over to the Chancellery for an interview with Hindenburg that was to prove fateful for himself, for Germany, and for the rest of the world. From a window in the Kaiserhof, Goebbels, Roehm and other Nazi chiefs kept an anxious watch on the door of the Chancellery, where the Fuehrer would shortly be coming out. ‘We would see from his face whether he had succeeded or not’, Goebbels noted. For even then they were not quite sure. ‘Our hearts are torn back and forth between doubt, hope, joy and discouragement’, Goebbels jotted down in his diary. ‘We have been disappointed too often for us to believe wholeheartedly in the great miracle.”

A few moments later they witnessed the miracle. The man with the Charlie Chaplin moustache, who had been a down-and-out tramp in Vienna in his youth, an unknown soldier of World War I, a derelict in Munich in the first grim postwar days, the somewhat comical leader of the Beer Hall Putsch, this spell-binder who was not even German but Austrian, and who was only forty-three years old, had just been administered the oath as Chancellor of the German Reich.

He drove the hundred yards to the Kaiserhof and was soon with his old cronies, Goebbels, Goering, Roehm and the other Brownshirts who had helped him along the rocky, brawling path to power. ‘He says nothing, and all of us say nothing’, Goebbels recorded, ‘but his eyes are full of tears.’

That evening from dusk until far past midnight the delirious Nazi stormtroopers marched in a massive torchlight parade to celebrate the victory. By the tens of thousands, they emerged in disciplined columns from the depths of the Tiergarten, passed under the triumphal arch of the Brandenburg Gate and down the Wilhelmstrasse, their bands blaring the old martial airs to the thunderous beating of the drums, their voices bawling the new Horst Wessel song and other tunes that were as old as Germany, their jackboots beating a mighty rhythm on the pavement, their torches held high and forming a ribbon of flame that illuminated the night and kindled the hurrahs of the onlookers massed on the sidewalks. From a window in the palace Hindenburg looked down upon the marching throng, beating time to the military marches with his cane, apparently pleased that at last he had picked a Chancellor who could arouse the people in a traditionally German way. Whether the old man, in his dotage, had any inkling of what he had unleashed that day is doubtful. …

A stone's throw down the Wilhelmstrasse Adolf Hitler stood at an open window of the Chancellery, beside himself with excitement and joy, dancing up and down, jerking his arm up continually in the Nazi salute, smiling and laughing until his eyes were again full of tears.

Shirer's account, which takes up a full page in the paperback edition, is completely given over to the dramatic nature of the event. Like Bullock, he mentions the crowd; but he adds to it a description of the Nazi stormtroopers which itself takes up one full paragraph and which is packed with details of their movements and activities. Again like Bullock, Shirer draws a contrast (which he calls a ‘miracle’, as opposed to Bullock's more judicious ‘improbable’) between Hitler's present position and his former life. But again his description of this one item takes up a whole paragraph, he lists four of Hitler's lowly roles as opposed to the three mentioned by Bullock, and he adds in three further contrasts (dealing with Hitler's age and appearance) for good measure.

So much for where Shirer covers the same ground as Bullock. But Shirer adds three quite new ingredients as well. First, he comments on the fateful nature of that day and in a three-part climax (italicized) looks forward briefly to the terrible consequences that were to stem from it. Second, he describes the reaction of Hitler's immediate entourage and quotes from Goebbels' own account of the event. And finally, in yet another three-part climax (also italicized), he focuses upon the mood of Hitler himself, dancing about at the open window and giving Nazi salutes. Shirer's recipe, so much more elaborate than the basic Bullock, produces a narrative which satisfies the reader's need for the authentic and the dramatic, and whets his appetite for more of the same in other episodes.

Now which of these two accounts would have appealed most to the reader in ancient Rome? Cicero has left us a record of what he expected from history, and it goes like this. ‘The narrative should be built up in both subject-matter and style. … A mere chronicle of events holds very little interest for us, no more than the entries in the official almanac. But the uncertain and varied fortunes of a man who rises to prominence provoke admiration, suspense, joy, sorrow, hope, or fear; and if they end in a striking dénouement, the mind enjoys one of the greatest pleasures that reading can give. … For nothing is more calculated to please the reader than the changes and chances of fortune.’³ Cicero, it is clear, is looking for an expanded and dramatic narrative; and it is Shirer who provides this, rather than Bullock.

At this point it is interesting to reflect why Shirer's account is so different from that of Bullock. Bullock is a professional historian, concerned with analysis and interpretation. This is clear as early as his title-page, where we read: Hitler: A Study in Tyranny. There can be few less dramatic words in English than the word ‘study’; and the success which Bullock's book has enjoyed is perhaps due to the fact that its distinguished author has risen above the limitations of his professional genre. For it has been argued that many modern historians are losing touch with the art of communication. Peter Gay, himself a well-known academic historian, discusses the whole problem in his book Style in History, and what he says is this:⁴

Much historical work is innocent of even a nodding acquaintance with the writer's art. We have all encountered those dreary, dutiful chronicles piling up mounds of facts that everyone knows or nobody wants to know; those narrow, earnest monographs choking in their garlands of ibids and parched in their deserts of charts. We have wondered at those mountainous and learned theses that do strive for distinction but founder in literary incoherence, with their style borrowed from the ungainly informativeness of the railway timetable. Whatever else it may be, history is not an art all of the time.

As a general rule, a modern historian has only a very limited readership. Provided he can persuade someone to publish his work, he is thereafter concerned only with the relatively few other professionals who will be competent to understand the issues he discusses. (The same is of course even more true of classicists. …)

By contrast, Shirer is a professional news reporter, and reporters have to satisfy a vast, non-specialist audience whose demands are twofold. On the one hand readers of newspapers demand consistent subject-matter and content; on the other, they demand that newspapers justify their name by presenting their repetitious material in new or different ways. The reader's appetite for news is particularly obvious during wartime, something with which the ancient Romans were obsessed: ‘the vicissitudes of battles’, writes Tacitus, ‘take over and stimulate the readers' imagination’.⁵ So too during the American Civil War a New York newspaper sold five times its normal circulation when it ran details of a big battle;⁶ and during the First World War one of Lord Northcliffe's editors wrote that ‘war not only creates a supply of news, but a demand for it. So deeprooted is the fascination in war and all things pertaining to it … that a paper has only to be able to put up on its placard, A Great Battle for sales to mount up.’⁷ It is appropriate to reflect that the fascination of war is still very much with us, as the opinion polls indicated during the Falklands campaign. Hardly a season goes by without there being at least one war-serial on television; and feature-length war films are still being produced. A Bridge Too Far, which cost twice as much as the actual operation it depicts,⁸ is only one of the more recent examples. Reviewing that film in The Times, a critic wrote that ‘audiences flock to war films to exult in scenes of battle, to identify with acts of courage, and vicariously share in military glory’.⁹ This is the kind of demand which Shirer, as an American reporter in Europe before and during the last war, was obliged to satisfy almost daily. And it is clear from the extract I quoted that his instincts as a reporter did not desert him when he came to write his book on Hitler's Germany. As with Bullock, this is clear as early as Shirer's title-page. When we read the words The Rise and Fall of the Third Reich we infer that the book will be concerned with the violent fluctuations of fortune which are beloved by the drama-hungry readers of newspapers and which were equally beloved, as we have seen, by Cicero and Tacitus.

It is, however, interesting to observe that Shirer's book has not found favour with professional historians. Reviewing the book some years after its publication, an American historian wrote as follows:

William Shirer's history of the Third Reich is woefully inadequate. … The inadequacies … could be dismissed out of hand if his book had not found an enormous audience. … The book is a literary tour de force. Much that is trivial has been elaborated because it is entertaining; much that is important has been omitted because it might be dull. … [The reading] is neither painful nor dull. … The lesson steadily mounts in interest to become increasingly exciting. The narrative, poor though it may be as history, has a sustained dramatic tension. Each episode moves to a climax, and the successive episodes achieve a grand climax—the Wehrmacht's defeat and Hitler's suicide.¹⁰

In this review we see neatly encapsulated the antithesis which has grown up between literature-and-entertainment on the one hand, and ‘real’ history on the other.

This antithesis can be illustrated again and again in books which deal with history or historiography. The well-known American reporter Theodore H. White begins his autobiography In Search of History by contrasting good reporters, who organize facts in ‘stories’, with good historians, who organize lives and episodes in ‘arguments’.¹¹ But the contrast need not be between history and a literary genre. Discussing the relationship between history and film, Donald Watt, professor of international history at the London School of Economics, alludes to the two common propositions that ‘the historian's main concern is accuracy; the producer of film and television is concerned with entertainment. The unspoken premise of the first proposition is that to be accurate is to be dull. The unspoken premise of the opposed proposition is that to be entertaining it is necessary to distort or misrepresent.’¹² Yet entertaining the reader was precisely the aim of the ancient historian, who was writing for a non-professional audience whose only other narrative genre, apart from historical prose, was epic poetry. Ancient historians, in fact, were expected to entertain their readers in exactly the same way as modern reporters, and statements of this expectation are found in ancient authors from the fourth century b.c. to the fifth century a.d., including such distinguished authors as Cicero and Tacitus. For example, the Hellenistic historian Duris in a famous fragment criticized two of his predecessors on the grounds that they did not provide entertainment in their works. When Cicero describes his own reading of the Greek historians, he says he usually reads them only for entertainment; and in his well-known letter to Lucceius, advising his friend how to write history, Cicero emphasizes its entertainment potential no less than nine times. The architect Vitruvius declared that the essence of history was to retain the reader's interest, and Pliny in his letters said that history provides entertainment no matter how it is written. We have already noticed that this was also the view of Tacitus.¹³

Now anyone who has read Phillip Knightley's brilliant book The First Casualty will know that some modern war reporters will stop at nothing, including fabrication, to entertain their readers. One example which Knightley happens not to mention concerns Hermann Goering, who, destined to be hanged on 20 October 1946, cheated the executioner by taking poison the previous day. This event greatly embarrassed certain reporters, who, never dreaming that Goering would commit suicide, had already despatched to their editors vivid accounts of his death by hanging.¹⁴

To see whether Roman historians, similarly dedicated to entertaining their readers, adopted similarly fallacious techniques, I should like to examine three passages:

The season was winter, and snow was falling in the area which lies between the Alps and the Apennines; there were marshes and rivers in the locality, intensifying the bitter cold. Moreover, men and horses had been turned out too quickly: there was no time to eat or snatch up coverings as protection against the cold. Warmth had drained from their bodies, and as they approached the vicinity of the river, the full force of the cold struck them more keenly.

As the campaigning season approached, Gaul still lay beneath its snow. The snow stretched … all across the hills as the army marched east from Cularo, a white blanket across the slopes of the Alps, embroidered by the black and leafless trees; the snow covered the gray ice of the frozen lakes, just beginning to show the seams and cracks of coming thaw. … Under the snow it is impossible to tell poor farm from rich farm, for snow forces farmers to shelter livestock and equipment indoors.

There were no tree trunks or roots by which a man could hoist himself up, only smooth ice and thawing snow, over which they were always rolling. … Four days were spent at the cliff, and the animals nearly perished of starvation; for the mountain tops are all practically bare, and such grass as does grow is buried beneath the snow. Lower down one comes to valleys, and slopes bathed in sunlight, and streams, and near to them are woods, and places more suitable for human habitation.

Which of these three passages was written by Livy? Unless you happen to know your Livy almost off by heart, the problem is not easy. Each passage deals with winter travel, each contains specific detail of an eyewitness nature, and each passage has a dramatic or romantic flavour about it. In fact the second of these passages was written by the American reporter Theodore H. White to whom I have already referred. It is taken from his book of Kennedy's presidential campaign in 1960 and I altered some geographical and other items in order to make my point. The original is:¹⁵

As the primary approached, Wisconsin still lay beneath its snow. The snow stretched … all across the hills as the plane flew west from New York, a white blanket across the slopes of the Appalachians, embroidered by the black and leafless trees; the snow covered the gray ice of the frozen lakes, just beginning to show the seams and cracks of coming thaw. … Under the snow it is impossible to tell poor farm from rich farm, for snow forces farmers to shelter automobiles and equipment indoors.

Now the radical journalist Paul Foot has written a highly critical review of White's book; and commenting on the passage I have just quoted, Foot says this:¹⁶

No detail is too irrelevant to include in this exhilarating chronicle; for, however banal, each detail proved that the historian himself was actually there. He too moved with the gods, and he could bring any literate American right into the Kennedy compound at Hyannisport or the cabin of the Kennedy campaign plane as it travelled from cheering crowd to cheering crowd, over the white snow, the blue sea, the green valleys, or whatever.

Theodore White is obsessed with descriptive detail: Foot interprets this obsession as White's proof that he actually took part in the events which he relates. No doubt Foot is quite correct; but we must also remember that descriptive detail is the hallmark of the fabricator of history. A classic case is that of Sir Edmund Backhouse, who spent most of his life in China and there forged the sources for the history of China on which he collaborated with an unsuspecting colleague and which for many years was accepted as the standard history of its period. As Trevor-Roper writes in his life of Backhouse, one element in the man's plausibility was his extraordinary circumstantiality and his minute and scrupulous detail. ‘Backhouse's language was never loose or vague. He impressed everyone by his remarkable “memory”, which in turn seemed a guarantee of truth.’¹⁷

Livy is the author of the two other passages: they are taken from his account of Hannibal's march across the Alps in Book 21.¹⁸ Livy's descriptions here are just as detailed as Theodore White's, yet Livy was writing two full centuries after the event itself had taken place. How did he come by such authentic description? It might have come from his sources, like the Greek historian Polybius, who wrote:¹⁹

The summits of the Alps, and the parts near the tops of the passes, are all quite treeless and bare, owing to the snow lying there continuously both winter and summer. But the parts half-way up on both sides are wooded and generally inhabitable.

This extract is indeed very similar to the second passage of Livy which I quoted. Polybius in his turn will have got such information from Silenus, who actually did accompany Hannibal across the Alps in 218 b.c. It is possible, in other words, to argue that Livy's description is ultimately based on an eyewitness account. But you will have noticed that Polybius mentioned only woods lower down. Livy mentioned woods too, but added valleys, streams, and slopes bathed in sunlight. His Latin for this last phrase is apricos colles: apricus is a favourite word of the poets, found in romantic and descriptive passages of Virgil and Horace,²⁰ but only here in the whole of Livy that has survived. I think that in places like this Livy is relying on no source at all but is adding purely fictitious detail which he remembered from his days at school when he would have been taught the basic techniques of how to describe landscapes.

Higher education at Rome was in the hands of professional rhetoricians whose main aim was the entertainment of their audience. ‘I say many things’, admitted one of them, ‘not because they please me but because they please my audience.’²¹ And we happen to know that one of the most enjoyable things for an audience to hear was the description of natural scenery or landscape. There is of course nothing unusual in this. John Buchan once observed that only two ingredients are essential to the successful thriller writer: an ability to create likeable characters, and a feel for landscape by which to satisfy the armchair traveller's hunger for a sense of place.²² And so we learn from numerous Roman authors that rhetoricians were expected to have a stock of ready-mixed, easy-to-use landscapes which they could insert into any composition in order to entertain their audience.²³

Now we have seen that Cicero believed that history should entertain its readers; but when in the 50s b.c. he looked back on the writings of the early Roman historians, he was extremely disappointed: they had, he complained, paid no attention to the description of landscape, giving only the barest indications of natural scenery.²⁴ Now history is of course a most appropriate genre in which to include the description of landscape; and since such descriptions were, as we have seen, extremely popular with audiences and readers in general, Cicero suggested that historians should make far more use of landscape description than hitherto. ‘The nature of historical writing’, he said, ‘demands chronological sequence and landscape description.’ In fact he even went so far as to define history as the genre ‘in which the narrative is elaborate and landscape descriptions are given’.²⁵

It seems clear from the extract I quoted above that Livy took Cicero at his word. But Roman historians were hardly great travellers, doing on-the-spot research into the distant places they described in their works. They were notoriously armchair writers, whose only course of action, if they took Cicero's advice (and they did), was to follow the standard but ever-popular methods of describing landscapes that were already being practised by rhetoricians. This is exactly what Livy has done in the modest passage which I quoted. Feeling that the information he had inherited from Polybius was a little dull, he filled it out with some descriptive detail. The detail sounds authentic enough, otherwise I would not have been able to compare it with the extract from Theodore White's book; but we have no guarantee at all that the detail is true. On the contrary, its utterly conventional nature and its similarity to descriptive passages elsewhere indicate that it is highly suspect.

Now it may be felt that this conclusion is relatively unimportant. Why should it matter if Livy's description on this occasion was invented rather than true? I think there are two reasons why it matters. First, there is evidence that the procedure recommended by Cicero and put into practice by Livy gained increasing popularity during the first century a.d., reaching its climax when Tacitus began to write. It is, I believe, significant that writers at the end of the first century a.d. refer to landscape description as a primarily historical device and not the rhetorical device which it had been for Cicero a century and a half earlier.²⁶ An excellent example is to be found in Curtius Rufus' History of Alexander the Great, written in the middle of the first century a.d. He is describing Persia:²⁷

At the foot of the mountains a spacious plain slopes down, a fertile land, abounding in many villages and cities. Through these fields the River Araxes rolls the waters of many torrents into the Medus … a river which is more favourable to the growth of vegetation than any other, and it clothes with flowers whatever it flows near. Plane trees and poplars cover its bank, so that to those who view them from far off, the groves along the banks seem to be a continuation of those on the mountains. For the shaded stream flows in a channel sunk deep in the soil, and over it hang hills which themselves are rich in foliage.

Another most effective and detailed piece of description. But Curtius was surely relying on the ignorance of his readers—much as did the makers of an early newsreel film which purported to show gallant British medical staff at work during the Boer War but which was actually shot on Hampstead Heath.²⁸ For the two alleged rivers are one and the same (there seems never to have been a river Medus) and the mountains in the area are barren and, to judge from present-day evidence, never supported extensive vegetation at all.²⁹ Curtius' description is simply a collection of various landscape motifs which he knew would paint a pretty picture for the pleasure of his readers. Cicero and Livy between them encouraged a fashion from which no later historian was immune.³⁰

There is a second reason why I think that the procedure adopted by Livy matters. If Livy could resort to invention on the occasion we have been discussing, there is nothing to stop him doing exactly the same thing on other, much more important, occasions. Livy is our main source for many of the great battles of Roman history: have we any guarantee at all that his descriptions of the scenes and sites of these battles are correct? How would we feel if a modern historian, having given a vivid account of the Battle of the Ardennes, later admitted that he knew little about the area except that it was covered with forest? In fact, if the procedure of the Roman historians is as I have described, it is perhaps creative writers, not historians, that rather come to mind. When setting out as a young man for America, W. H. Davies wrote a full description of the country before his ship left Liverpool dock; and what he had done was discovered only because he posted his account from the ship's first port of call in Ireland, in a letter bearing a British stamp. Similarly Anthony Burgess advised that ‘it's best to imagine your own foreign country. I wrote a very good account of Paris before I ever went there. Better than the real thing.’³¹ But whereas such practices cause no concern in the cases of Davies and Burgess, we would be scandalized if they were adopted by modern historians and we should be equally alarmed when we find that they are adopted on a large scale by their ancient counterparts.

Of course many people do not think much of Livy as a historian; and Curtius Rufus is generally acknowledged to be a second-rate writer of whom nothing better could be expected. Tacitus, the greatest of the Roman historians, is the yardstick against which to judge. So let us turn to the third book of Tacitus' Histories, where he describes an incident which allegedly took place in the civil wars of a.d. 68-9:³²

I find that some widely read historians vouch for the truth of the following story. The victors displayed such disregard for right and wrong that a trooper, claiming to have killed his brother in the recent battle, demanded a reward from his leaders. Common morality deterred them from honouring the murder, and the very nature of civil war from punishing it. In the end, it seems, they decided to put the man off by saying that the reward he deserved was too great to be paid on the spot. There is no further information. However, according to Sisenna, an equally ghastly act had occurred in a previous civil war, for in the battle against Cinna on the Janiculum, a soldier of Pompeius Strabo killed his brother, and then, when he realised what he had done, committed suicide. Thus in earlier generations merit evoked keener appreciation, and wicked actions keener remorse. Anyhow, it will not be inappropriate for me to cite these and similar anecdotes from ancient history when the context calls for lessons in right conduct or consolation for evil.

The extract is typically Tacitean. It deals with a subject, namely fratricide, which was nearly as popular as landscape with Roman audiences,³³ and you can almost hear Tacitus' disappointment when he says that he has no further information about it. But this admission turns out to be a device whereby he can introduce a corresponding example from a previous civil war in order to contrast earlier nobility with present decadence. Now the example from the previous civil war is attested not only by Sisenna, as Tacitus says, but also by Livy;³⁴ but what of the incident which Tacitus himself describes? Though he claims rather defensively to rely on ‘some widely read historians’, did the incident ever take place? There are two reasons why I am sceptical.

Several chapters earlier in the same book of the Histories, Tacitus had described another incident with remarkable similarities to the present one:³⁵

The slaughter was even more noteworthy because of an instance of parricide. I record the incident and the names on the authority of Vipstanus Messalla. A recruit to the Hurricane Legion, one Julius Mansuetus from Spain, had left a young lad at home. Soon after, the boy came of age, and having been called up by Galba for service in the Seventh, chanced to encounter his father in this battle and wounded him seriously. As he was searching the prostrate and semiconscious figure, father and son recognised each other. Embracing the dying man, the son prayed in words choked by tears that his father's spirit would be appeased and not bear him ill-will as a parricide: the act was not a personal one; and what was one soldier but an infinitesimal fraction of the forces engaged in the civil war? With these words, he took up the body, dug a grave, and discharged the last duty to his father. Some nearby troops noticed this, then more and more; and so throughout the lines ran a current of wonder and complaint, and men cursed this cruellest of all wars. Yet this did not stop them killing and robbing relatives, kinsmen and brothers; they told each other that a crime had been done, and in the same breath did it themselves.

Parricide was another subject which the Romans enjoyed,³⁶ and Tacitus' enthusiasm for this incident is clear both from his opening words that the slaughter was ‘even more noteworthy’ because of it, and also from the number of emotional words he uses in the course of the passage (e.g. oblatum forte, semianimem s exsanguem, agnitus agnoscensque, uoce flebili, miraculum, questus, saeuissimi belli execratio). Now the details of this episode are solemnly repeated as true by the author of a recent book on this period,³⁷ but as long ago as 1918 the French scholar Courbaud realized that something was amiss: ‘Tacitus, who hardly ever cites any source, has taken care (wisely, in the circumstances) to name the authority on whom he is based. But did things take place exactly as he describes? Did the son bury his father there and then in the middle of the battle? Are his complaints not stereotyped … ? We recognize the technique: Tacitus has worked up the details with which he was furnished, and developed them so that they appeal more to our sense of drama and pathos.’³⁸

Courbaud is on the right lines, and the very fact that this incident so closely resembles the previous one suggests to me that one or other of them is unlikely to be genuine. But my second piece of evidence will, I think, cast doubt on the genuineness of both of the episodes which Tacitus describes. Among the small collection of poems which are attributed to the younger Seneca, but which are in fact anonymous and of unknown date, there are two pieces of verse which purport to describe an incident in the famous civil war between Octavian and Mark Antony which took place after the civil war which Sisenna and Livy described but before that which Tacitus describes in the Histories. The first of these poems begins as follows:³⁹

A soldier from Octavian's side called Maevius dared to jump on an enemy boat and overturn it; but his happiness depended upon a personal loss: he was fated to be a sacrilegious victor since he was unwittingly exultant over the murder of his brother. While he was tearing away his enemy's armour in the hunt for battle trophies, he recognised his brother's sad face. His noble act had been a crime.

After almost a dozen lines of platitudinous soliloquy, in the course of which Maevius decides to commit suicide, the poem concludes as follows:

He was still uncertain whose sword to use for the deed. ‘Shall I die by my own sword, stained as it is by an unspeakable murder? The person for whom I am dying will supply the sword for my death.’ With these words he took up his brother's sword and fell on top of his brother, one and the same hand putting an end to both victor and victim.

The second poem is shorter and less complicated, but otherwise very similar. It begins thus:⁴⁰

Maevius was delighted to think he had killed one of the enemy, but his happiness turned to sadness since it was his brother he had killed. He could not escape finding out: while he savagely stripped the body of his victim, he came upon trophies that belonged to his own family. Simultaneously he recognized his brother and his crime.

This poem too concludes with a soliloquy in which Maevius decides to redeem himself by committing suicide.

Both these poems are desperately impoverished, as my translations make only too clear, but they nevertheless provide valuable evidence that what we are dealing with is a stock motif. Fratricide, as I have said, was a conventional and popular theme; and Maevius is a stock name.⁴¹ What the authors of these poems have done is take the incident in the first civil war, to which Tacitus referred at the end of the first extract I quoted, and applied it to the civil war between Octavian and Mark Antony. And in my opinion Tacitus himself has adopted a very similar procedure, except that, being an infinitely superior writer, he has accomplished his task more effectively. In the second passage of his which I quoted, he has in effect told the Maevius story and applied it to the civil wars of a.d. 68-9; he has used an almost identical recognition scene, but changed the crime from fratricide to parricide and substituted a noble burial for the more melodramatic suicide. In the first passage of his which I quoted, Tacitus has rung further changes: he has retained the crime of fratricide but made the murderer claim credit for his deed, and to emphasize still further the twist he has introduced, he then retells the original story of fratricide and suicide on which, in my opinion, all the other versions are based.

You may of course object that history does tend to repeat itself, often in the most curious details. From Robert Graves's Goodbye to All That we learn that in 1914 a place called Festubert became a nightmare when the inmates of the local lunatic asylum, caught in cross-fire between the German and British lines, broke out from their enclosure and ran all over the countryside.⁴² And during Operation Market Garden thirty years later, as we learn from Cornelius Ryan, German ammunition caches exploded in woods near a place called Wolfheze, and a mental institute received direct hits—whereupon sixty terrified inmates, mostly women, started to wander about the woods.⁴³ A more bizarre example, reputedly the greatest series of coincidences in history, is associated with the Menai Straits. On three occasions (in 1664, 1785, and 1860) a ship has sunk there with all passengers except one. On each occasion the date of the disaster was 5 December, and on each occasion the name of the sole survivor was Hugh Williams.⁴⁴ Certainly in the first of these examples, and perhaps in the second, no one would wish to deny the genuineness of the incidents on the grounds that something very similar was recorded elsewhere. But of course the repetition of motif from one author to another is commonly used nowadays to deny the authenticity of incidents described in ancient poetry. Horace, we are told, cannot really have thrown away his shield at the battle of Philippi, as he alleges, because several of the Greek lyric poets had already told similar stories of themselves.⁴⁵ This argument has recently been subjected to sarcastic criticism by an American scholar, who points out that in their respective autobiographies Mortimer Wheeler, A. J. Ayer, and E. R. Dodds each describe an encounter with a prostitute (a different one in each case, of course!) whose charms each of them rejected.⁴⁶ Are we to accuse these distinguished scholars of plagiarism and fabrication? Presumably not. But that is merely to underline the difference between ancient and modern. The kind of operation which I have been conducting on isolated passages of Livy and Tacitus can be repeated time and again on the texts of these and other ancient historians. Recently I was able to demonstrate that one of the most celebrated episodes in the whole of Tacitus' Annals is largely fabrication by the historian,⁴⁷ a conclusion which the latest commentator on the Annals has accepted, rightly adding that it has troublesome implications.⁴⁸

There are two obvious conclusions which can, I think, be drawn from my argument. The first is simply that ancient and modern historiography are two quite different things. Because we use the same word ‘history’ to describe what is now written by A. J. P. Taylor and what was once written by Tacitus, for example, we tend to imagine a continuous tradition of historiography in which the difference between these two authors is merely one of chronology. We tend, in other words, to approach both authors, ancient and modern, with the same assumptions. Yet nothing could be more dangerous. What we ought to be doing is approaching ancient historians as the writers of literature which they are. They should be compared with Latin poets, as I have suggested in the case of Tacitus, or with modern reporters or creative writers, as I have suggested in the cases of Livy and Curtius respectively. Now if the ancient historians were indeed writers of the kind I have described, it follows that their works cannot simply be used as historical evidence by modern historians in the traditional way. Although epigraphy, archaeology, and numismatics naturally contribute to our knowledge of ancient history, it is upon the texts of the ancient historians themselves that most of our information about the detailed events of the ancient world traditionally depends. But if I am right to suggest that these texts are different in kind from what they are generally assumed to be, then the study of ancient history itself requires modification. Our primary response to the texts of the ancient historians should be literary rather than historical since the nature of the texts themselves is literary. Only when literary analysis has been carried out can we begin to use these texts as evidence for history; and by that time, as I hope I have indicated, such analysis will have revealed that there is precious little historical evidence left. The implications of this are indeed troublesome.

The second conclusion to be drawn from my argument follows on from the first, and it is this. One cannot conduct the kind of scholarly activity which I am describing without an appropriate knowledge of Greek and Latin. To think that one can carry out literary analysis on the basis of a Penguin translation and achieve the potentially significant results which I have mentioned is simply to live in a fool's paradise. Now it is no secret that in the early 1970s the number of candidates for A-level Latin suddenly and sharply declined, with the result that there have been many fewer students applying to university to study Latin. This situation has not, however, been caused by any unpopularity of Latin at school level, nor is it due to some endemic deficiency in the language and literature of Rome which it has taken 2,000 years to detect. Latin has been the victim of circumstances. The early 1970s saw the realization of an educational system which, as I understand it, was intended among other things to provide more pupils with a wider choice of subject than previously. As far as Latin and some other subjects are concerned, the theory has gone badly wrong. Either because it was regarded as an ‘élitist’ subject and so out of step with the allegedly egalitarian atmosphere of the times, or simply because the timetable difficulties of the new schools excluded minority subjects from the curriculum, Latin has been almost squeezed out. Schoolchildren are being deprived of the opportunity of knowing the language and literature of a civilization which has greatly influenced Western European society. […]

‘War is the father of all things’, said the Greek philosopher Heraclitus,⁴⁹ and war is the subject to which most historians in the ancient world devoted their attention. In this lecture I have suggested that these writers are different in kind from their modern counterparts and are more appropriately compared with writers in other genres, both ancient and modern. By way of a final illustration of my thesis, I should like to leave you with the following quotation. It is the opening paragraph to a longer piece:

This is an episode rich in disasters. … Four successive rulers have died violent deaths … and the country fallen victim to catastrophes which are either without precedent or have not occurred for centuries: whole towns burned down … and the capital severely damaged by fires which destroyed its most venerable buildings. … There has been adultery in high places. The whole Mediterranean swarmed with refugees, and the cliffs of its rocky islands ran with blood. … But the episode is not without its compensating features. … People's relatives stood resolute and loyal even under torture … and prominent men driven to suicide faced their last agony with unflinching courage.

Violence, bloodshed, death, torture, acts of individual heroism, and sex. These are the traditional elements of sensational war literature, and the paragraph might very well have introduced almost any Middle Eastern coup d'état as reported in a Sunday tabloid during the past decade or more. In fact, however, the paragraph was written early in the second century a.d. by the most respected and allegedly most sober of all the Roman historians: Tacitus. It is the opening paragraph of the work we know as the Histories,⁵⁰ and it was written to whet the appetite of his readers. It seems to have had the desired effect, since Pliny on reading it wrote to his friend Tacitus predicting that the Histories would be immortal.⁵¹ The prediction, as we have seen this evening, has turned out to be true—which is perhaps more than can be said for the Histories themselves.

Notes

1. Hitler: A Study in Tyranny (1975 edn.), 250.

2. The Rise and Fall of the Third Reich (1964 edn.), 16-17.

3. Cic. De Or. 2. 63, Fam. 5. 12. 5 and 4.

4. Style in History (1975), 186-7.

5. Tac. A. 4. 33. 3.

6. Knightley (1975), 23.

7. Knightley (1975), 85.

8. According to The Sunday Times, 23 October 1977, p. 12.

9. Philip French, The Times, 24 June 1977, p. 9.

10. W. O. Shanahan, American Historical Review, 68 (1962-3), 126-7.

11. In Search of History (1979), 2.

12. D. Watt, ‘History on the Public Screen I’, in P. Smith (ed.), The Historian and Film (1976), 169.

13. Duris fr. 1 Jacoby; Cic. De Or. 2. 59, Fam. 5. 12. 4-5; Vitr. 5 praef. 1; Plin. Ep. 5. 8. 4; Tac. A. 4. 33. 3.

14. Airey Neave, Nuremberg (1978), 316.

15. The Making of the President 1960 (1961), 82-3.

16. The Politics of Harold Wilson (1968), 12.

17. H. Trevor-Roper, A Hidden Life: The Enigma of Sir Edmund Backhouse (1976), 277.

18. Respectively 21. 54. 7-8 and 36. 7, 37. 4-5.

19. Polyb. 3. 55. 9.

20. Virg. Ecl. 9. 49, G. 2. 522, Aen. 5. 128, 6. 312; Hor. S. 1. 8. 15, C. 1. 8. 3, 26. 7, 3. 18. 2, Epist. 1. 14. 30, AP 162.

21. Cestius Pius ap. Sen. Contr. 9. 6. 12.

22. According to R. W. Winks, The Historian as Detective: Essays on Evidence (1968), 3.

23. See Sen. Contr. 2 praef. 3, 2. 1. 13; Quint. 4. 3. 12; Brink on Hor. AP 16-18, Woodman on Vell. 96. 3. On the transferability of such topoi (conventional motifs) from one composition to another see e.g. Cic. Inu. 2. 48 and Sen. Contr. 1 praef. 23.

24. Cic. De Or. 2. 53 ‘sine ullis ornamentis monumenta solum … locorum … reliquerunt’.

25. Respectively De Or. 2. 63 ‘rerum ratio ordinem temporum desiderat, regionum descriptionem’ and Or. 66 ‘(historia) in qua et narratur ornate et regio saepe … describitur’.

26. Cf. Plin. Ep. 2. 5. 5 ‘descriptiones locorum … non historice tantum sed prope poetice prosequi fas est’; Quint. 10. 1. 33 ‘licet tamen nobis in digressionibus uti uel historico nonnumquam nitore’.

27. Curt. 5. 4. 6-8.

28. See Knightley (1975), 75.

29. For the Araxes and the non-existent Medus see RE ii. 404 s.v. Araxes 4; the Medum flumen at Hor. C. 2. 9. 21 is the Euphrates (see Nisbet-Hubbard ad loc.), which flows away from the Araxes at right angles. The lack of vegetation in the area is observed by J. C. Rolfe in his edition (Loeb, 1946) ad loc. (I acknowledge the help of Prof. O. A. W. Dilke with some of the points in this note.) [A more positive view of Curtius' passage, however, is now taken by J. E. Atkinson in his recent commentary (1994) ad loc.]

30. See also Sall. J. 17-19, H. 3. 61-80; R. Syme, Sallust (1964), 194-5; Thomas (1982), 2-5.

31. For these two examples see P. Fussell, Abroad (1980), 174-5.

32. Tac. H. 3. 51 (trans. K. Wellesley, slightly changed).

33. See Winterbottom (1974), index III s.v. ‘parricide’ (there loosely defined as ‘killing of father or other near relative’).

34. Liv. per. 79.

35. Tac. H. 3. 25. 2-3 (trans. K. Wellesley, slightly changed).

36. See n. 34.

37. K. Wellesley, The Long Year a.d. 69 (1975), 148.

38. E. Courbaud, Les Procédés d'art de Tacite dans les ‘Histories’ (1918), 153-4.

39. [Sen.] Epig. 69 Prato, lines 13-36. For help with the interpretation of this and the following epigram I am much indebted to Prof. R. H. Martin and to my wife Dorothy.

40. [Sen.] Epig. 70 Prato, lines 3-18.

41. Or so I inferred, perhaps over-optimistically, from Hor. Epod. 10; yet Fraenkel (1957), 26 n. 3, notes that Mevius, used as a symbolic name (like Titus, etc.), occurs frequently in legal texts. [And see now Mankin on Hor. Epod. 10. 2.] Prof. W. Liebeschuetz has pointed out to me that the parricide-motif recurs in Sil. 9. 66-177, a passage discussed by K.-H. Niemann, Die Darstellung der röm. Niederlagen in den Punica des Silius Italicus (1975), 174-7 (a reference I owe to Dr W. R. Barnes). Niemann observes that an earlier scholar (R. Rebischke, De Silii Italici Orationibus (1913), 103-4) had suggested that ‘unwitting parricide in civil war’ was perhaps a suasorial theme in the rhetorical schools. [See also Spaltenstein on Sil. 9. 66.]

42. Goodbye to All That (1966 edn.), 146.

43. A Bridge Too Far (1977 edn.), 185. Another very similar incident had taken place in northern France slightly earlier (see J. Keegan, Six Armies in Normandy (1982), 185), and, amazingly, Dr G. A. Loud has pointed out yet another example from near Lille in 1940 (see A. Bryant, The Turn of the Tide 1939-1943 (1957), 127).

44. See D. Wallenchinsky and I. and A. Wallace, The Book of Lists (1977 edn.), 463. On the idea of history repeating itself see G. W. Trompf, The Idea of Historical Recurrence in Western Thought (1979).

45. Hor. C. 2. 7, with Nisbet-Hubbard ad loc. for the evidence from Greek lyric. Add B. Seidensticker, ‘Archilochus and Odysseus’, GRBS 19 (1978), 5-22.

46. W. M. Calder III, ‘The Spurned Doxy: An Unnoticed Topos in English Academic Autobiography’, CW 73 (1980), 305-7.

47. A. 1. 61-5 is ‘borrowed’ from H. 2. 70 and 5. 14-15: see Woodman (1979) [= Ch. 5 below].

48. See Goodyear (1981), 108.

49. Heraclitus 53 DK = 29 Markovich.

50. Tac. H. 1. 2. 1-3. 1.

51. Plin. Ep. 7. 33. 1, presumably referring to the first parts of the Histories to be ‘published’.

This is the text (slightly changed in places) of an inaugural lecture originally entitled ‘From Hannibal to Hitler: The Literature of War’ and delivered at the University of Leeds on Monday 31 January 1983.

Abbreviations

Tacitus' works are generally abbreviated as A.: (Annals), H.: (Histories), Agr.: (Agricola), G.: (Germania), and D.: (Dialogus). When referring to passages of the Annals, however, I often omit A. altogether if the reference is clear from the context; likewise I omit the book number when repeated reference is made to the same book of the Annals.

Periodical abbreviations are generally as in L'Année philologique; abbreviated references to other secondary works may be traced via the following list and the Bibliography.

ANRW: W. Haase and H. Temporini (eds.), Aufstieg und Niedergang der römischen Welt (Berlin and New York, 1972-).

BGU: Berliner griechische Urkunden (Berlin, 1895-).

CAH: ² A. K. Bowman, E. Champlin and A. Lintott (eds.), Cambridge Ancient History, vol. x (2nd. edn.; Cambridge, 1996).

CE: Carmina Latina Epigraphica, ed. F. Buecheler and E. Lommatsch, 3 vols. [= Anthologia Latina II] (Leipzig, 1895-1926).

CHCL: P. E. Easterling and E. J. Kenney (eds.), Cambridge History of Classical Literature, 2 vols. (Cambridge, 1982-5).

CIG: Corpus Inscriptionum Graecarum.

CIL: Corpus Inscriptionum Latinarum.

EJ V. Ehrenberg and A. H. M. Jones, Documents Illustrating the Reigns of Augustus and Tiberius (2nd edn.; repr. Oxford, 1976).

FGrH: Die Fragmente der griechischen Historiker, ed. F. Jacoby (Berlin and Leiden, 1923-58).

K-S: R. Kühner and C. Stegmann, Ausführliche Grammatik der lateinischen Sprache, ii. Satzlehre, Parts 1 and 2 (4th edn.; repr. Hanover, 1971).

LH: C. S. Kraus and A. J. Woodman, Latin Historians (Greece & Rome: New Surveys in the Classics, 27) (Oxford, 1997).

L-H-S: M. Leumann, J. B. Hofmann, and A. Szantyr, Lateinische Grammatik, ii. Syntax und Stilistik (revised edn., Munich, 1972).

OCD: Oxford Classical Dictionary.

OLD: Oxford Latin Dictionary.

RAC: Reallexikon für Antike und Christentum.

RE: Paulys Real-Encyclopädie der classischen Altertumswissenschaft.

RICH: A. J. Woodman, Rhetoric in Classical Historiography (London, Sydney, and Portland, Or., 1988).

SCPP: Senatus Consultum de Cn. Pisone Patre.

SEG: Supplementum Epigraphicum Graecum (Leiden, 1923-).

TLL: Thesaurus Linguae Latinae.

Works Cited

Fraenkel, E. (1957), Horace (Oxford).

Knightley, P. (1975), The First Casualty: From the Crimea to Vietnam (London).

Winterbottom, M. (1974), The Elder Seneca, Loeb edn. (Cambridge, Mass. and London).